Before we start, as always, this tab entry is purely a product of personal opinion, watching much comedy central, and is solidly based on the wild conspiracy theory (if this isn't dual noise, what is?) It is completely unrelated to any real-world company or university, including the ones I worked for in the past or present. On the other hand, it is well and truly dedicated to every Rodney Dangerfield in O.R. practice (you belong to this club only if you already knew that).
In academia, a professor would be happy to get their work published in a leading peer-reviewed O.R journal and often, this acceptance defines the degree of success or failure of a research project. Almost all new ideas in a university make it to some journal or conference, but very few practical innovations in the industry gain visibility. They are either patented, or remain hidden as a company's intellectual property. At the end of the day, the Edelmans, like the Oscars, are as much (and probably more) a tribute to an organization's upper management for being O.R. friendly, as it is to the guys in the trenches who actually pull off the O.R. innovations.
So how does the common O.R. guy in the company gain peer-recognition? (suggestions most welcome here :-)
After prototyping is done, what does the practitioner do? Sit around tooling, while waiting for analytical support calls ? There are a few choices (or few choices if you have a bad boss), depending on when your next project starts. You can try to patent the most practical approach. After all, in practice, the proof is in the pudding. Or you can publish a non-proprietary version of your findings. A cynic may say that what this really means is that you released an 'unpractical' version, but this not the case - at least not always. Patenting is common place in the IT world, but a relatively less popular option in the O.R. world. A patent in your resume can make you look more 'result-oriented'. A future employer may be worried if he/she observes way too many publications during your past job ("so did u do any real work this millennium?").
On the other hand, if you plan to move on to academia after a while, publishing is not such a bad idea. And it's not bad for a future industry job as well, since it serves as a solid, peer-reviewed reference for your scientific skills. It's also a good marketing and recruiting tool, since your company gains recognition in the scientific community as a place that promotes cutting edge R&D. A good manager would recognize these benefits. A common fear that is unfounded is that publishing = giving away your intellectual property. As long as you keep your engineering ideas out of scope, my own experience is that it can be counter-productive for a competitor to try and directly reproduce a product from a journal paper that appears in print 2-4 years after the idea was implemented in a product (or not). Add to that a couple of years that it takes to go from idea to finished product, and you get the picture. Then you realize that you would have been better off building something for today's customer on your own, rather than relying on recycled ideas.
Patenting versus publishing is a personal choice. Acads are shocked to see frivolous patents. But for every seemingly hideous patent, there exists a new publication for yet another factor-of-2 approximation algorithm for an NP-Hard problem that can be readily 'solved' in milliseconds for real data instances. The answer to the original question is not clear cut. And what about presenting at conferences?
Nowadays, companies count the number of patent applications filed by their R&D group. If such a metric is employed, perhaps patenting should be the first option. This scenario is more likely if you work in a non-traditional O.R. industry, and in such a situation, presenting at a conference can provide visibility and is a nice trade-off. Conference presentations takes up much less time than publishing, and you can get your ideas out there quickly. On the other hand, conference proceedings seem to be valued less in our field. Perhaps practitioners can present more of their work at a dedicated conference. These should be a conference of practitioners, and organized by, and for the benefit of practitioners. That would be wicked. Why the Gettysburg clause you ask? When profs meet, its an conference, but when practitioners and engineers meet, its labeled a workshop. As we all know, when the rocket stays up, it's hailed a scientific success, and if not, its an engineering failure. There's still no respect for 'R&D'.
Tuesday, March 16, 2010
Tuesday, February 23, 2010
O.R. Ball - Defense and Offense
Often times, the difference between the theory and practice of O.R. is the same as that between a carefully choreographed brain surgery at Boston General versus brain repair performed at the MASH 4077. In the former case, the idea is come back with a generic breakthrough that can be safely re-used for all surgeries. In the latter case, the aim is to get through the Korean war (on TV albeit) with the least number of casualties among the huge number of wounded that randomly show up.
The textbook approach prides itself on being data-agnostic and coming up with new "small polynomial" techniques that exploit special structure in a problem class. Data agnosticism is something we practitioners can ill-afford since the proof is in the pudding. If we build the best recipe that is optimized for no more than 6-7 guests, and we never expect to see more than ten guests, ever, then it is pointless to fuss over a complex recipe-generator that optimally serves a thousand guests, but can't serve five as quickly and as well. And if there are thousands of such five-guest parties to be served, the former approach wins hands down. Of course, now if we were optimally designing a dam or two, things would be a little different since strict service levels come into play. So it is case dependent, and bringing your skill and art into play to deal with such differences and putting your money where your model is, happens to be one of the reasons why O.R. practice is never dull.
Textbooks recommend that we exploit problem structure. Industrial optimization exploits problem structure to an extent, but can and should exploit the structure in data. The Simplex method that is making such a robust comeback via GUROBI is worst-case exponential and rarely does a bad job because good implementations thrive on real-world LP data and the numerical properties of computers. In the real world, even strongly NP-hard optimization problems are quite manageable. Do not let textbooks scare you! After all, there were no computers around when architects in South India optimally designed the Brihadeeswara temple a thousand years ago - the design required that the tall temple's shadow be constrained to within its (convex) perimeter any time of day. They achieved an elegant and 'feasible' design that stands 216 feet high while using incredibly heavy granite stones to do this. And yes, the location of the symbolic 'idol' of the deity coincides with the centroid of the overall structure. Now this has got to be one great place for spiritual reflection or thinking O.R!
It's a strange dichotomy in the global O.R. community. The universities are focused on 'defense' and abhor seeing any "2^n" kind of numbers anywhere - a systematic O.R. approach that generates lower bounds to protect your answers, but is also capable of forcing turnovers, i.e., can be quickly converted into good quality solutions with a little bit of imagination. The book on the Traveling Salesman Problem by the stalwarts of our discipline is quite fascinating in this context. The O.R. business community goes with whatever brings home the Dosa, exponential or otherwise, and there is healthy scorn for the 'defense' part. Approximation / local optimum / randomized methods can be thought of as being part of the 'offense'. It gets the glory and can be used to quickly initiate a product, and that perhaps is a reason why many a young practitioner is hooked to it. However, it is generally a good idea to consider using both approaches. Over the product life cycle, it is usually the best defense-offense combo that wins the game, all other things being equal.
The textbook approach prides itself on being data-agnostic and coming up with new "small polynomial" techniques that exploit special structure in a problem class. Data agnosticism is something we practitioners can ill-afford since the proof is in the pudding. If we build the best recipe that is optimized for no more than 6-7 guests, and we never expect to see more than ten guests, ever, then it is pointless to fuss over a complex recipe-generator that optimally serves a thousand guests, but can't serve five as quickly and as well. And if there are thousands of such five-guest parties to be served, the former approach wins hands down. Of course, now if we were optimally designing a dam or two, things would be a little different since strict service levels come into play. So it is case dependent, and bringing your skill and art into play to deal with such differences and putting your money where your model is, happens to be one of the reasons why O.R. practice is never dull.
Textbooks recommend that we exploit problem structure. Industrial optimization exploits problem structure to an extent, but can and should exploit the structure in data. The Simplex method that is making such a robust comeback via GUROBI is worst-case exponential and rarely does a bad job because good implementations thrive on real-world LP data and the numerical properties of computers. In the real world, even strongly NP-hard optimization problems are quite manageable. Do not let textbooks scare you! After all, there were no computers around when architects in South India optimally designed the Brihadeeswara temple a thousand years ago - the design required that the tall temple's shadow be constrained to within its (convex) perimeter any time of day. They achieved an elegant and 'feasible' design that stands 216 feet high while using incredibly heavy granite stones to do this. And yes, the location of the symbolic 'idol' of the deity coincides with the centroid of the overall structure. Now this has got to be one great place for spiritual reflection or thinking O.R!
It's a strange dichotomy in the global O.R. community. The universities are focused on 'defense' and abhor seeing any "2^n" kind of numbers anywhere - a systematic O.R. approach that generates lower bounds to protect your answers, but is also capable of forcing turnovers, i.e., can be quickly converted into good quality solutions with a little bit of imagination. The book on the Traveling Salesman Problem by the stalwarts of our discipline is quite fascinating in this context. The O.R. business community goes with whatever brings home the Dosa, exponential or otherwise, and there is healthy scorn for the 'defense' part. Approximation / local optimum / randomized methods can be thought of as being part of the 'offense'. It gets the glory and can be used to quickly initiate a product, and that perhaps is a reason why many a young practitioner is hooked to it. However, it is generally a good idea to consider using both approaches. Over the product life cycle, it is usually the best defense-offense combo that wins the game, all other things being equal.
Tuesday, February 9, 2010
Blog Review: The Age of Analytics
Web-tooling on the Monday after Superbowl for analytic content, I found this great new blog 'The Analytic Age' for O.R. practitioners and enthusiasts alike.
About the author - Dr. V. Ramakrishnan, has a strong background in O.R. and has a distinguished track record as an analytics entrepreneur, among other things. He currently teaches at the MIT Sloan School, and was VP-Chief Scientist at Profit Logic (and later at Oracle Retail), a company which designed and built the hugely successful Markdown Optimization Product, and was acquired by Oracle a few years ago.
The first commentary on analytics in the blog provides valuable real-world insight into the practice of this craft. It is encouraging to learn that businesses are looking for believable and optimized decision analytic tools. The kind of 'A' that is not just about how cool the technology inside is, or how good it looks on the outside, or leaves you stranded with an impressive array of numbers and summary reports; rather, the kind of 'A' that provides reliable prescriptive answers based on sound business sense. These business requirements directly play to the strength of O.R.
There's magic, and then there is practical magic. The latter is more likely to bring the audiences back.
About the author - Dr. V. Ramakrishnan, has a strong background in O.R. and has a distinguished track record as an analytics entrepreneur, among other things. He currently teaches at the MIT Sloan School, and was VP-Chief Scientist at Profit Logic (and later at Oracle Retail), a company which designed and built the hugely successful Markdown Optimization Product, and was acquired by Oracle a few years ago.
The first commentary on analytics in the blog provides valuable real-world insight into the practice of this craft. It is encouraging to learn that businesses are looking for believable and optimized decision analytic tools. The kind of 'A' that is not just about how cool the technology inside is, or how good it looks on the outside, or leaves you stranded with an impressive array of numbers and summary reports; rather, the kind of 'A' that provides reliable prescriptive answers based on sound business sense. These business requirements directly play to the strength of O.R.
There's magic, and then there is practical magic. The latter is more likely to bring the audiences back.
Wednesday, January 20, 2010
The iPod and the O.R. Designer
The web is full of business articles that extol the benefits of a great, overall user experience that has come to define the iPod product. These articles roughly state that we knowingly accepted a discrete approximation of continuous music and enjoyed the experience because it was delivered in a really convenient manner. Business commentators even go so far as to say that features don't matter any more. Not surprisingly, there are huge lessons to be learned for those us who make a living by building apps with 'O.R. inside' for business end-users.
If your OR program in school was part of the Industrial and Systems Engineering (ISE) department, then we would recognize that these are not novel ideas. We have seen this before in our ISE labs and it resembles 'Human Factors' engineering. It was always exciting to know what the HFE guys were up to - they were generally designing all this ergonomic stuff - keyboards, wheelchairs, etc. We O.R. grads were optimizing big-deal industrial problems with sophisticated math and did not pay enough attention to such "low-tech and qualitative" HFE ideas. Designing for an optimal business-user experience can be thought of as an area that combines OR with HFE. First a look at a partial checklist of design questions that we can relate to.
Does your O.R. app:
- require the user to have an O.R. PhD to operate?
- feel paranoid about the sophistication of the O.R. technology inside?
- obsess about run time with little regard for quality?
- have bewildering layers of menus filled with dials for 'costs', 'penalties', etc.?
- change answers wildly with small changes in input?
- come back with a blank stare if it encounters infeasibility somewhere inside?
...
You get the idea. An O.R. practitioner not only has to worry about the stuff that is inside - traditionally, we've been good analysts and trained for that, but it's about time we become equally good designers - few, if any, O.R. graduate programs teach that. Perhaps we should be teaming up with our next door HFE neighbors on this. As we continue to explore this theme, we will notice interesting connections between what's inside and outside the app.
If your OR program in school was part of the Industrial and Systems Engineering (ISE) department, then we would recognize that these are not novel ideas. We have seen this before in our ISE labs and it resembles 'Human Factors' engineering. It was always exciting to know what the HFE guys were up to - they were generally designing all this ergonomic stuff - keyboards, wheelchairs, etc. We O.R. grads were optimizing big-deal industrial problems with sophisticated math and did not pay enough attention to such "low-tech and qualitative" HFE ideas. Designing for an optimal business-user experience can be thought of as an area that combines OR with HFE. First a look at a partial checklist of design questions that we can relate to.
Does your O.R. app:
- require the user to have an O.R. PhD to operate?
- feel paranoid about the sophistication of the O.R. technology inside?
- obsess about run time with little regard for quality?
- have bewildering layers of menus filled with dials for 'costs', 'penalties', etc.?
- change answers wildly with small changes in input?
- come back with a blank stare if it encounters infeasibility somewhere inside?
...
You get the idea. An O.R. practitioner not only has to worry about the stuff that is inside - traditionally, we've been good analysts and trained for that, but it's about time we become equally good designers - few, if any, O.R. graduate programs teach that. Perhaps we should be teaming up with our next door HFE neighbors on this. As we continue to explore this theme, we will notice interesting connections between what's inside and outside the app.
Saturday, January 16, 2010
NRF-2010: Some thoughts
The annual convention for retailers ended this week. It was a jam-packed convention with all the big players making their presence felt. The first thing I noticed is that the retail industry is used to the notion of 'analytics', which for most part is tied to forecasting and making good use of Terra bytes of data. Nevertheless, as a retailer looks to maximize their revenue dollars and improve operational efficiency, O.R. has an increasingly large role to play. Retail O.R. problems are pretty challenging in both scale and complexity and given that retail margins are small, improved decision making for such complex planning and operational problems using 'decision science' can make a big difference. Textbook retail decision models provide a starting point for building innovative practical solutions, even if they cannot be directly deployed within products, and the retail world abounds with hidden, interesting O.R. problems if you start looking for them.
A good introductory (non-technical) book that retail analytic experts have suggested as part of recommended reading is "Retailing Management" by M. Levy and B. Weitz.
A good introductory (non-technical) book that retail analytic experts have suggested as part of recommended reading is "Retailing Management" by M. Levy and B. Weitz.
Saturday, January 9, 2010
99th National Retail Federation Expo, NYC, Jan 10-13, 2010
Every player in the retail arena will be at this flagship NRF event for the retain industry. I hope to catch the latest buzz in retail O.R. and fathom customers appetite for O.R. methods, and tab about it when I get back. Retail is still a relatively young field when it comes to O.R., but its like an iceberg. Most of the cool O.R. stuff is in products rather than journals. While journals provide valuable insight into the kind of problems faced by retailers, I haven't seen too many solutions presented in the literature as yet that consistently work well in practice, although I expect to see this trend reverse in the near future.
Wednesday, January 6, 2010
Murder Down Under - Where is the data?
The recent spate of attacks on Indian students in Australia has created serious diplomatic tension between the two countries. On one hand, we have the (apparently) hysterical English-language media of India that magnifies every such report. On the other hand, we have the Australian police and media apparently living in denial. Where does the truth lie?
A bit of history first. India and Australia have often had a love-hate relationship. Indian troops fought and died alongside ANZACs during WW2, yet Australia refused to admit Indian immigrants for decades after the conflict, even preferring those from 'white' Italy (against whom they had just fought against). In this context, their recent apology for the 'lost innocents and forgotten Australians' also comes to mind. In recent times, on a less serious note, the Australian and Indian cricket teams have been involved in a few, well-publicised, on-field verbal and legal clashes over the last 3 years. On the other hand, the biggest fans of Australian cricket are probably in India. Steve Waugh the former Australian cricket captain, runs a reputed charity organization in Kolkata, India and is much loved and admired. And of course, after 9/11 and the Patriot act, increasing numbers of Indian students who usually head for the U.S. instead choose to study down under.
Not much statistics is available on what's really going on. In the last year, an estimated 80-100, 000 Indian students studied at universities down under making it a really lucrative business for the Aussies. A total of 105 violent attacks have been documented (some fatal). Today, we had the Indian envoy say that "most of these attacks were opportunistic", even as the Indian government issued a travel warning to students there. At the current rate, slightly more than 1 in 1000 Indian students have been victims. It would be interesting to compare this rate to the overall violent-crime rate there, as well as the rate for a comparable ethnic population (China has a larger number of students there and may be a good candidate). In particular, overall crime rates have been on the increase in Melbourne. Per wikipedia " ... From July 2008 to July 2009 assaults reportedly increased 8.7% in the city of Melbourne, 4.3% in the city of Yarra, 12.5% in the city of Port Phillip 17% in the city of Stonington ... "
At this point, its not clear if these vicious thugs hate Indians (I'm sure that not all Indian students are like Chatur Ramalingam) or just hate everybody they meet. Hopefully, a comprehensive statistical analysis will succeed where denial and hysteria has failed.
Jan 10 update: To be sure, no student deserves to be assaulted, even if you are a Chatur.R.
A bit of history first. India and Australia have often had a love-hate relationship. Indian troops fought and died alongside ANZACs during WW2, yet Australia refused to admit Indian immigrants for decades after the conflict, even preferring those from 'white' Italy (against whom they had just fought against). In this context, their recent apology for the 'lost innocents and forgotten Australians' also comes to mind. In recent times, on a less serious note, the Australian and Indian cricket teams have been involved in a few, well-publicised, on-field verbal and legal clashes over the last 3 years. On the other hand, the biggest fans of Australian cricket are probably in India. Steve Waugh the former Australian cricket captain, runs a reputed charity organization in Kolkata, India and is much loved and admired. And of course, after 9/11 and the Patriot act, increasing numbers of Indian students who usually head for the U.S. instead choose to study down under.
Not much statistics is available on what's really going on. In the last year, an estimated 80-100, 000 Indian students studied at universities down under making it a really lucrative business for the Aussies. A total of 105 violent attacks have been documented (some fatal). Today, we had the Indian envoy say that "most of these attacks were opportunistic", even as the Indian government issued a travel warning to students there. At the current rate, slightly more than 1 in 1000 Indian students have been victims. It would be interesting to compare this rate to the overall violent-crime rate there, as well as the rate for a comparable ethnic population (China has a larger number of students there and may be a good candidate). In particular, overall crime rates have been on the increase in Melbourne. Per wikipedia " ... From July 2008 to July 2009 assaults reportedly increased 8.7% in the city of Melbourne, 4.3% in the city of Yarra, 12.5% in the city of Port Phillip 17% in the city of Stonington ... "
At this point, its not clear if these vicious thugs hate Indians (I'm sure that not all Indian students are like Chatur Ramalingam) or just hate everybody they meet. Hopefully, a comprehensive statistical analysis will succeed where denial and hysteria has failed.
Jan 10 update: To be sure, no student deserves to be assaulted, even if you are a Chatur.R.
Friday, January 1, 2010
The Three Idiots and O.R
The three most enjoyable Hindi films of 2009 were released in December, with '3 idiots' from last week being the best of the lot. It's a great way to start the new year with a post on this movie. More so because it is about engineering, and about having a passion for building something new and better, something which we O.R' ers take a good deal of pride in.
It's an entertaining cry for a serious overhaul of education methods in India. As much as Obama wants the U.S. to match India and China in science and math, one hope it's done the right way. Many of the Indians and Chinese who have made their mark as technology and business leaders have done so despite the system. The pitfalls of rote learning (a trademark of Indian education) is brought out hilariously in the movie. The 'rote learning' pesky teacher's pet, aptly named Chatur (clever) Ram-alingam, does everything 'in memory', and is brilliantly essayed by native Californian Omi Vaidya (he plays 'Sadiq' in the TV sitcom 'the Office'). Chatur is from Uganda and completely memorizes his keynote eulogy of a tyrannical professor Dr. Viru.S at the annual university function, given that he is Hindi-illiterate. However, the three idiots manage to do a 'find and replace' in the draft of the English transliteration of the speech that substitutes "Chamat-kari" (miracle worker) with "Balaat-kari" (serial molester), and the resulting chaos sets up the rest of the movie.
On a more serious note, the movie reminds us of the high student-suicide rate in India, and argues that underemployment, even if it means a higher pay in a job totally unrelated to your field of expertise, is undesirable, and job satisfaction is far more important. It would be interesting to survey the level of job satisfaction for O.R practitioners. I personally feel that it will high. Why? Nobody can be forced to choose a career in "Operations Research". That's a plus point of having such a weird name. Secondly, an O.R graduate can work in a bank, in a social organization, or study ant colonies and apply his skills and feel at home. O.R. blesses you with the science of how to improve processes, almost independent of the context. Successful ideas in airlines and logistics are now being applied to retail analytics, medical science, and green technologies.
O.R. does and should continue to reward creativity and innovation, and hopefully in 2010, each of us ' O.R. idiots' who practice this field will build or improve one new, innovative product with 'O.R inside' (and progress beyond rote memorization of the KKT conditions).
The movie should be legally available on YouTube in a couple of months, and I hope the English subtitles will not be done by a rote-learner!
It's an entertaining cry for a serious overhaul of education methods in India. As much as Obama wants the U.S. to match India and China in science and math, one hope it's done the right way. Many of the Indians and Chinese who have made their mark as technology and business leaders have done so despite the system. The pitfalls of rote learning (a trademark of Indian education) is brought out hilariously in the movie. The 'rote learning' pesky teacher's pet, aptly named Chatur (clever) Ram-alingam, does everything 'in memory', and is brilliantly essayed by native Californian Omi Vaidya (he plays 'Sadiq' in the TV sitcom 'the Office'). Chatur is from Uganda and completely memorizes his keynote eulogy of a tyrannical professor Dr. Viru.S at the annual university function, given that he is Hindi-illiterate. However, the three idiots manage to do a 'find and replace' in the draft of the English transliteration of the speech that substitutes "Chamat-kari" (miracle worker) with "Balaat-kari" (serial molester), and the resulting chaos sets up the rest of the movie.
On a more serious note, the movie reminds us of the high student-suicide rate in India, and argues that underemployment, even if it means a higher pay in a job totally unrelated to your field of expertise, is undesirable, and job satisfaction is far more important. It would be interesting to survey the level of job satisfaction for O.R practitioners. I personally feel that it will high. Why? Nobody can be forced to choose a career in "Operations Research". That's a plus point of having such a weird name. Secondly, an O.R graduate can work in a bank, in a social organization, or study ant colonies and apply his skills and feel at home. O.R. blesses you with the science of how to improve processes, almost independent of the context. Successful ideas in airlines and logistics are now being applied to retail analytics, medical science, and green technologies.
O.R. does and should continue to reward creativity and innovation, and hopefully in 2010, each of us ' O.R. idiots' who practice this field will build or improve one new, innovative product with 'O.R inside' (and progress beyond rote memorization of the KKT conditions).
The movie should be legally available on YouTube in a couple of months, and I hope the English subtitles will not be done by a rote-learner!
Friday, December 11, 2009
Analytics in Sport - Cricket and Operations Research - 1
Analytics in Cricket is not a new idea. A lot of OR folks, especially from the so-called Commonwealth nations, would be pleasantly surprised to learn that OR has been an integral part of cricket (in particular, the limited overs versions) for the last 16-odd years. This is because of the official induction of the Duckworth-Lewis rules for weather-interrupted matches into the rulebook. Mr. Duckworth and Mr. Lewis are OR/Math guys. According to cricinfo, the latter is/was the chairman of the western branch of the Operational Research society in the U.K. See this old article in ORMS Today on their work.
Cricket is more than a hundred years old, and is the second-most followed sport on this planet (thanks to more than a billion and a half cricket-mad fans from the Indian subcontinent, including this author). India is the No.1 test cricket team in the world today (after 77 years of hapless performance), and cricket is now big business that is growing in size, what with all the professional leagues like the IPL springing up. Among all sport, cricket mirrors life the most, and its rules suitably reflect this. It's best been described as having a cleverly disguised gentlemanly exterior which hides a series of fierce one-on-one gladiatorial contests of blood, guts, and stamina, which is then wrapped in a chess-game of wits and strategy, and tied up with strings of psychological tactics of 'mental disintegration'.
While D/L rules prescribe revised targets for rain-affected limited-over matches, these rules can also be used as a decision-support tool for teams to figure out an optimal trajectory to achieve the target during a run-chase. More generally, the run-chase problem can also be formulated as (stochastic) Dynamic Programming problem. Constrained resources include batsmen available (10), and overs available (20 or 50, depending on the whether its a T20 game or a 50-over game), with objective being to get at least one run more than the opposition before either of these resources are exhausted. At any stage of the game, the teams can tailor their tactics according to this optimal trajectory that can be recalculated after after every ball or over.
Stastical tools that analyze batting and bowling performances, and for stuff like SWOT analysis, have regularly been a part of cricket in recent times, much like baseball. A new idea proposed here is to analytically decide on how to make optimal use of the newly introduced review system, similar to the challenge system in tennis and perhaps the NFL. Statistically, cricket umpires tend to get 1 in 10 decisions wrong (very impressive given that theres 10 or more ways of getting out in cricket :-). At the highest skill-levels of cricket, i.e., country- versus-country test cricket that is played over 5 days, a bad decision can result in the team at the receiving end of this decision, getting pummelled into a defensive position for a couple of days under the hot sun.
It is quite likely that a probability model can be built around this idea. For example, given that there are 10 wickets available in an inning, using a geometric probability model,
prob(at least one error) = 1 - prob (getting all 10 right) = 1 - (0.9^10) = 0.65.
So there is roughly a 2/3 chance that the average umpire will get at least one decision wrong in a completed inning and this number can increase if the particular umpires in that match are known to be more error-prone. Given that a team is allowed only two (wrong) reviews, the idea is to come up with a reviewing strategy that ensures that you only challenge decisions that maximizes your team's expected advantage-level for the remainder of the game. If you use these challenges frivolously, you are left with no recourse later in an inning. If you do not use it at all, you are more than likely to suffer at least one bad decision per inning.
Interestingly, the international cricket council (ICC), which is the governing body of cricket states that with the review system in place, the statistical error-rate has improved to roughly 95%. This means that the probability of an error in a complete inning is reduced from 65% to 40%, implying that it is statistically more likely that there will no wrong decision in an inning with the review system in place that allows an on-field umpire to change a decision based on evidence from video and audio footage (upon request from either team).
Tennis players at grand-slam events tend to invoke this challenge during key times of the match ('big points'), where you are at a 'cliff'. e.g., at break-point. In the past we have seen great players (McEnroe at a certain French Open comes to mind), who've mentally lost matches from winning positions because of what they perceived to be an unjust call. However, doing so only at these points may not be optimal, since a highly-probable bad call at 30-30 by a line-judge is a good candidate for review. In other words, there is both a probability value as well as a consequence-cost (risk/reward) associate with this decision.
In most situations, these decisions have to made on-the-spot, with limited external-feedback available. Rather than rely wholly on instinct and emotions ("I'm sure I am not out"), it would be nice for the player under the spotlight to have some analytic ammo to go with bat or ball.
Having said all this, the great teams (like the ones from Caribbean in the 80s) and great players like the late Donald Bradman (who is considered to be the greatest Australian who ever lived), and currently, Sachin Tendulkar of India (who is a shoo-in for the greatest living Indian today), are the ones who raise their game to a higher-level when faced with adversity, and overcome such bad calls. On the other hand, a little bit of O.R could make this task a little less difficult.
- correction added: D/L rules have been used for nearly 16 years now.
Cricket is more than a hundred years old, and is the second-most followed sport on this planet (thanks to more than a billion and a half cricket-mad fans from the Indian subcontinent, including this author). India is the No.1 test cricket team in the world today (after 77 years of hapless performance), and cricket is now big business that is growing in size, what with all the professional leagues like the IPL springing up. Among all sport, cricket mirrors life the most, and its rules suitably reflect this. It's best been described as having a cleverly disguised gentlemanly exterior which hides a series of fierce one-on-one gladiatorial contests of blood, guts, and stamina, which is then wrapped in a chess-game of wits and strategy, and tied up with strings of psychological tactics of 'mental disintegration'.
While D/L rules prescribe revised targets for rain-affected limited-over matches, these rules can also be used as a decision-support tool for teams to figure out an optimal trajectory to achieve the target during a run-chase. More generally, the run-chase problem can also be formulated as (stochastic) Dynamic Programming problem. Constrained resources include batsmen available (10), and overs available (20 or 50, depending on the whether its a T20 game or a 50-over game), with objective being to get at least one run more than the opposition before either of these resources are exhausted. At any stage of the game, the teams can tailor their tactics according to this optimal trajectory that can be recalculated after after every ball or over.
Stastical tools that analyze batting and bowling performances, and for stuff like SWOT analysis, have regularly been a part of cricket in recent times, much like baseball. A new idea proposed here is to analytically decide on how to make optimal use of the newly introduced review system, similar to the challenge system in tennis and perhaps the NFL. Statistically, cricket umpires tend to get 1 in 10 decisions wrong (very impressive given that theres 10 or more ways of getting out in cricket :-). At the highest skill-levels of cricket, i.e., country- versus-country test cricket that is played over 5 days, a bad decision can result in the team at the receiving end of this decision, getting pummelled into a defensive position for a couple of days under the hot sun.
It is quite likely that a probability model can be built around this idea. For example, given that there are 10 wickets available in an inning, using a geometric probability model,
prob(at least one error) = 1 - prob (getting all 10 right) = 1 - (0.9^10) = 0.65.
So there is roughly a 2/3 chance that the average umpire will get at least one decision wrong in a completed inning and this number can increase if the particular umpires in that match are known to be more error-prone. Given that a team is allowed only two (wrong) reviews, the idea is to come up with a reviewing strategy that ensures that you only challenge decisions that maximizes your team's expected advantage-level for the remainder of the game. If you use these challenges frivolously, you are left with no recourse later in an inning. If you do not use it at all, you are more than likely to suffer at least one bad decision per inning.
Interestingly, the international cricket council (ICC), which is the governing body of cricket states that with the review system in place, the statistical error-rate has improved to roughly 95%. This means that the probability of an error in a complete inning is reduced from 65% to 40%, implying that it is statistically more likely that there will no wrong decision in an inning with the review system in place that allows an on-field umpire to change a decision based on evidence from video and audio footage (upon request from either team).
Tennis players at grand-slam events tend to invoke this challenge during key times of the match ('big points'), where you are at a 'cliff'. e.g., at break-point. In the past we have seen great players (McEnroe at a certain French Open comes to mind), who've mentally lost matches from winning positions because of what they perceived to be an unjust call. However, doing so only at these points may not be optimal, since a highly-probable bad call at 30-30 by a line-judge is a good candidate for review. In other words, there is both a probability value as well as a consequence-cost (risk/reward) associate with this decision.
In most situations, these decisions have to made on-the-spot, with limited external-feedback available. Rather than rely wholly on instinct and emotions ("I'm sure I am not out"), it would be nice for the player under the spotlight to have some analytic ammo to go with bat or ball.
Having said all this, the great teams (like the ones from Caribbean in the 80s) and great players like the late Donald Bradman (who is considered to be the greatest Australian who ever lived), and currently, Sachin Tendulkar of India (who is a shoo-in for the greatest living Indian today), are the ones who raise their game to a higher-level when faced with adversity, and overcome such bad calls. On the other hand, a little bit of O.R could make this task a little less difficult.
- correction added: D/L rules have been used for nearly 16 years now.
Monday, November 30, 2009
O.R. Practice: Doomed by Success?
Did you hear about the dissident O.R practitioner who was sentenced to 'a death by a 1000 cuts' ? Legend has it that his body was found remarkably intact, integral ....
The claim here is that an commercial O.R solution to a real-life problem has a finite shelf-life. The graph of potential improvement for a product is concave, and follows the law of diminishing returns. Most of our recent posts have focused on the need to ensure that the first solution has the 'O.R. inside' stamp, since entrenched heuristics of unknown quality are surprisingly resistant to replacement by more smart-logic based O.R methods.
But what happens after you have something with O. R inside? As a former colleague's professor asked him 'So do you sit around waiting for the model to break?' The answer is sometimes yes. Other times, our twitching O.R genes compel us to keep improving upon the solution and after a while, the effort is not worth the improvement. The better the prior effort was, the less likely that you will be allowed to tinker with it any further. Pretty soon it's set in stone and it just becomes an automaton. After a while, it may even cease to be of much competitive value to a company and the functions are likely to be outsourced to a cookie-cutter vendor.
A car designer can spend an entire career endlessly tweaking cars, but an O.R practitioner has to diversify and cannot expect to retire with the same company by endlessly tweaking a product that she or he created and cherished much. O.R. is such a nebulous and ill-defined field in practice that your next manager or director may not have clue as to what the heck your field is, let alone what it is that you have been doing so far. Without strong backing from the highest levels within the management ("Edelman VPs"), the best O.R. efforts can come to nought or go straight to conference and we, the practitioners, have to move on to a different job.
Anyway, this is just one person's take. It would be instructive to hear the experiences of other practitioners.
The claim here is that an commercial O.R solution to a real-life problem has a finite shelf-life. The graph of potential improvement for a product is concave, and follows the law of diminishing returns. Most of our recent posts have focused on the need to ensure that the first solution has the 'O.R. inside' stamp, since entrenched heuristics of unknown quality are surprisingly resistant to replacement by more smart-logic based O.R methods.
But what happens after you have something with O. R inside? As a former colleague's professor asked him 'So do you sit around waiting for the model to break?' The answer is sometimes yes. Other times, our twitching O.R genes compel us to keep improving upon the solution and after a while, the effort is not worth the improvement. The better the prior effort was, the less likely that you will be allowed to tinker with it any further. Pretty soon it's set in stone and it just becomes an automaton. After a while, it may even cease to be of much competitive value to a company and the functions are likely to be outsourced to a cookie-cutter vendor.
A car designer can spend an entire career endlessly tweaking cars, but an O.R practitioner has to diversify and cannot expect to retire with the same company by endlessly tweaking a product that she or he created and cherished much. O.R. is such a nebulous and ill-defined field in practice that your next manager or director may not have clue as to what the heck your field is, let alone what it is that you have been doing so far. Without strong backing from the highest levels within the management ("Edelman VPs"), the best O.R. efforts can come to nought or go straight to conference and we, the practitioners, have to move on to a different job.
Anyway, this is just one person's take. It would be instructive to hear the experiences of other practitioners.
Wednesday, November 18, 2009
Theory of Inadvertent Cutting Planes and 2-D LSP
The perils of employing heuristics of unknown quality are often disregarded in practice, all in the interest of 'time to market' and 'practical' solutions for NP-Hard optimization problems. See, for example, Dr. Gerald Brown's papers and presentations along with the late Dr. Rick Rosenthal on this topic. (also see old post on 'the paradox of optimality'). Importantly, Dr. Brown reminds us of the huge difference between 'known unknowns' and 'unknown unknowns', before we start to make the poor assumption that NP-Hard automatically implies a quick, randomized heuristic approach. Dr. Michael Trick's recent blog entry on NP-Hardness is illuminating. Such heuristics do have a role to play in O.R. practice, depending on the business problem at hand. We attempt to illustrate, to the non-technical audience in particular, using a simple example:
The 2-D Laughing Stock Problem
PICTURE 1: shows the feasible region (a polygon), the optimal solution, and the one the heuristic algorithm found.
PICTURE 2: shows the new constraint added by the user that reduces the feasible space. The previous heuristic solution is infeasible now. Solver re-optimizes.
PICTURE 3: shows the new heuristic solution that is near-global optimal. The bewildering user experience so far is that he/she has added a highly restrictive constraint, yet the app ended up with a dramatically better solution, one even better than the "optimal". Imagine driving a car that has such heuristics built into its steering response.
The 2-D Laughing Stock Problem
PICTURE 1: shows the feasible region (a polygon), the optimal solution, and the one the heuristic algorithm found.
PICTURE 2: shows the new constraint added by the user that reduces the feasible space. The previous heuristic solution is infeasible now. Solver re-optimizes.
PICTURE 3: shows the new heuristic solution that is near-global optimal. The bewildering user experience so far is that he/she has added a highly restrictive constraint, yet the app ended up with a dramatically better solution, one even better than the "optimal". Imagine driving a car that has such heuristics built into its steering response.
Sunday, November 1, 2009
On decisioneering and dealing with sneering detractors
Part of an O.R practitioners job involves selling O.R to non-believers in the organization. Yet many of us in the O.R comfort-zone are firm non-believers that there even exist such non-believers. After all, isn't 'science of better' or its applied counterpart 'decisioneering' self-explanatory? It isn't. The 'analytics' bandwagon is going to ensure that. Last time we looked at the identity crisis facing the poor OR guy. Today, we'll examine more related aspects.
When we say a product has got 'O.R inside', what do we really mean? Is it because it's been autographed by that lost O.R scientist whose owlish ^oo^ spectacles always makes u think 'infinite loop', or, is it the bullet-proof C++ codes of O.R algorithms, the fiendishly reformulated optimization model, or the brand-new, low-latency, 16M$, 32-node, 64-bit, 128-GB SMP RAM parallel machine (yummy!) that smashes thru all your Lagrangian subproblems in a jiffy? or perhaps it's all in the GUROBI or CPLEX solvers that implements the fundamental algorithms?
The old bilateral debate of man v machine, in this context, starts with 'Math v Programming', and in true O.R fashion, cascades into some NP-complete combinatorial debate. heh. The obvious answer to many may be 'all the above', but called me biased - I feel that its the well-trained O.R grad, her/his model and solution approach that seals the deal here. Everything else is essentially a commodity, and can be quickly purchased, and therefore form the supporting cast (The real answer of course is 'none of the above'. It's the power point decks that made it all happen).
Seriously, a practitioner has to have all the soft skills to ensure that O.R gets some small share of credit in such projects, especially when things go right. After all, when its fails, its because of the O.R inside. It's because of you. Everything else was purchased and they work just fine! Suddenly, you alone know which constraint is hurting profits the most, or why a few more discrete variables kill run-times, or if the exponential service time assumption holds. Which brings me to probabilistic 'OR inside' models in practice (more on that another day). By design, its going to give you 'wrong' answers some of the time - unlike deterministic models that provide the illusion of correctness all the time. Good luck selling that!
When we say a product has got 'O.R inside', what do we really mean? Is it because it's been autographed by that lost O.R scientist whose owlish ^oo^ spectacles always makes u think 'infinite loop', or, is it the bullet-proof C++ codes of O.R algorithms, the fiendishly reformulated optimization model, or the brand-new, low-latency, 16M$, 32-node, 64-bit, 128-GB SMP RAM parallel machine (yummy!) that smashes thru all your Lagrangian subproblems in a jiffy? or perhaps it's all in the GUROBI or CPLEX solvers that implements the fundamental algorithms?
The old bilateral debate of man v machine, in this context, starts with 'Math v Programming', and in true O.R fashion, cascades into some NP-complete combinatorial debate. heh. The obvious answer to many may be 'all the above', but called me biased - I feel that its the well-trained O.R grad, her/his model and solution approach that seals the deal here. Everything else is essentially a commodity, and can be quickly purchased, and therefore form the supporting cast (The real answer of course is 'none of the above'. It's the power point decks that made it all happen).
Seriously, a practitioner has to have all the soft skills to ensure that O.R gets some small share of credit in such projects, especially when things go right. After all, when its fails, its because of the O.R inside. It's because of you. Everything else was purchased and they work just fine! Suddenly, you alone know which constraint is hurting profits the most, or why a few more discrete variables kill run-times, or if the exponential service time assumption holds. Which brings me to probabilistic 'OR inside' models in practice (more on that another day). By design, its going to give you 'wrong' answers some of the time - unlike deterministic models that provide the illusion of correctness all the time. Good luck selling that!
Friday, October 16, 2009
Identity crisis for the O.R practitioner
If you work in an industry that is saturated by O.R, then this is not for you. Familiarity tends to breed contempt there, and like a bad Steven Segal movie, your work goes straight to conference, heh. It's fun working in an area that is barely touched by O.R, especially if you are a new OR PhD. Your graduate advisor sent you off on your way last week with words like 'remember, no cuts, no glory'. You cant wait to get started ..
"O.R." You realize the name doesn't help. All those rumours in grad school were true! You just go with a simple 'decision science'. Three months into your job, you launch a satyagraha to get your basic tools like Gurobi to work with. You begin your first project.
First off, the sales and pre-sales folks (science is anathema to them but they bring home the bread that allow desk-jockeys like us to tool with OR, so no quarrel) ask you 'so if you are going to solve this using CPLEX, why do we need you? If you can explain 'reformulation', 'NP-Hard' to them and save your new job, your next conference talk will be a piece of cake.
Fact: CPLEX or Gurobi cannot solve any real-life problem directly. Skilled O.R People do (duh!). MS word is just as useful for that purpose.
Next, your strategy folks ("where powerless science meets power point") ask you: why cant our competitor also use CPLEX to solve these problems. whats the big value in decision science?
Not surprisingly, its a bit more difficult to convince folks in the stratosphere that there's real magic in O.R. Heck, it doesn't matter anyway, since they are going to forget it in a couple of days and get back to their ethereal kingdom.
Fact: These tools have blazing fast, industrial strength implementations of fundamental algorithms. The secret sauce is in your business-specific meta-models and meta-algorithms that is independent of the vendor that implements the fundamental tools used inside them.
Facts aside, the word 'Meta' convinces them that you are on to something. Next, you deal with the IT guys. They play for the home team. Problem: OR guys cant code, even though every one of us is convinced otherwise. Your prototype C++ program looks so random, they cant believe that something deterministic comes out of it. The name 'Math Programming' doesnt help either. To save the company from you, they place their trust in their beautifully coded 30-class, 30-line randomized algorithm that everybody now believes will do just as well and go with that. What does the customer care about optimality? It's a battle for another day. Right now, you are getting ready to present your work at a conference ...
Disclaimer: This is a work of pure OR-fiction. Except for CPLEX, Gurobi, and O.R, everything else in this tab has no resemblance to reality.
-------------------------------------------
Here on forward, the Tooler's Tab will waste time solely on OR and analytic topics. For more serious stuff like cricket, fictional detectives, and Indian music, follow the link to my blog on the right panel.
"O.R." You realize the name doesn't help. All those rumours in grad school were true! You just go with a simple 'decision science'. Three months into your job, you launch a satyagraha to get your basic tools like Gurobi to work with. You begin your first project.
First off, the sales and pre-sales folks (science is anathema to them but they bring home the bread that allow desk-jockeys like us to tool with OR, so no quarrel) ask you 'so if you are going to solve this using CPLEX, why do we need you? If you can explain 'reformulation', 'NP-Hard' to them and save your new job, your next conference talk will be a piece of cake.
Fact: CPLEX or Gurobi cannot solve any real-life problem directly. Skilled O.R People do (duh!). MS word is just as useful for that purpose.
Next, your strategy folks ("where powerless science meets power point") ask you: why cant our competitor also use CPLEX to solve these problems. whats the big value in decision science?
Not surprisingly, its a bit more difficult to convince folks in the stratosphere that there's real magic in O.R. Heck, it doesn't matter anyway, since they are going to forget it in a couple of days and get back to their ethereal kingdom.
Fact: These tools have blazing fast, industrial strength implementations of fundamental algorithms. The secret sauce is in your business-specific meta-models and meta-algorithms that is independent of the vendor that implements the fundamental tools used inside them.
Facts aside, the word 'Meta' convinces them that you are on to something. Next, you deal with the IT guys. They play for the home team. Problem: OR guys cant code, even though every one of us is convinced otherwise. Your prototype C++ program looks so random, they cant believe that something deterministic comes out of it. The name 'Math Programming' doesnt help either. To save the company from you, they place their trust in their beautifully coded 30-class, 30-line randomized algorithm that everybody now believes will do just as well and go with that. What does the customer care about optimality? It's a battle for another day. Right now, you are getting ready to present your work at a conference ...
Disclaimer: This is a work of pure OR-fiction. Except for CPLEX, Gurobi, and O.R, everything else in this tab has no resemblance to reality.
-------------------------------------------
Here on forward, the Tooler's Tab will waste time solely on OR and analytic topics. For more serious stuff like cricket, fictional detectives, and Indian music, follow the link to my blog on the right panel.
Subscribe to:
Posts (Atom)