THE BOOK--Playing The Percentages In Baseball

This is a non-event and has been anticipated for weeks. Woods was bound to drop after missing so much time, it’s the nature of the ranking system.

But that’s not important. The author has a HOF vote, so he must know what he’s talking about when it comes to baseball. Blurgh.

Devil’s Advocate: Everything looks like a nail when you have a hammer. The use of computers and accessibility to high levels of computational power shape how we think about things. We’ve seen it in the sciences and humanities, and of course it will effect sports analysis. We should have the humility to acknowledge that it’s not the only way, and not even the most right way, and it undoubtedly reflects the tools used and therefore reflects a bias.

The “computer” was programmed, in accordance to some specification (either real or imagined by the programmers). It’s just doing something a person can do, but, in this case, faster. The World Golf Rankings and BCS issues are not technical and not about computers at all. So what if the stupidity/crookedness/accuracy manifests in software?

The Forbester’s argument seems to be little more than blaming the messenger. It just so happens that the messenger in this case is Si-based instead of C-based.

they guy may have a point. i’m having a tough time resisting my Orwellian compulsions urging me to use a computer to read his blog instead of driving the the nearest store to purchase a stack of dead trees with his words printed on them.

Whilst posting #4, the BCS and its frequent criticisms occurred to me (as they also apparently did to Harry in #3—great minds think alike). It would be nice were there a web site which 1) spelled out in implicit detail how the BCS rankings are compiled every week, and 2) allowed site visitors to tweak criteria and see the resulting changes in the BCS rankings. Such a site would remove much of the perception that the BCS rankings come out of a black box. That’s the sort of perception that led to the Forbes article now at issue.

In #6, I meant “explicit” detail in the spelling-out business. I hate being stupid.

As a programmer and baseball freak I know you can make “rankings” to come out any # of ways, how many yrs have I spent trying to correlate OOTP Baseball ratings to actual stats they produce when sumulated!!!!

too many

http://www.pbrl.org for a league

The article really misrepresents Westwood’s tour results this year. It makes it seem like he hasn’t played much if at all.  He played in 11 events this year, with one win, 2 seconds, 6 top 10’s and 8 top 25’s. He also won the tidy sum of 3.4 million dollars.  That is one helluva year.  Golfers don’t get too many “counting stats” other than tourneys won, top 10 finishes, etc. (and I guess monies won).  Most golfers, even the great ones, not named Tiger won’t win more than a tourney or 2 per year.  A golfer who plays in 11 tournaments and finishes in the top 10 in 6 of them is A LOT better than a golfer who plays in 25 tourneys and hardly ever finished in the top 10 and never wins even one.  While Westwood may not be truly the best golfer in the world, his computer ranking is hardly absurd.  Not even close (to being absurd).  It is NOT like Palmeiro winning the gold glove despite playing 28 games in the field.  Does this writer even know anything about golf?

Alan
The computers are only one third of the bcs. wikipedia explains how it is compiled but in brief of the six programs the highest and lowest score for each school is thrown out. The colley programs web site does allow you to change results and see the effect but that is unusual. Of the 6 programs only colley and Wolfe are explained enough to allow me to try and replicate them although there is some info on the Massey and sagarin programs but billingsley and anderson are black boxes as far as I can see. The other problem is that only wins and losses but not margin of victory are taken into account by the computers by order of the bcs but the human polls clearly take mov into account and this is often the reason why the computers disagree with the humans.

However at least the computers take all the games into account whereas the pollsters cannot have watched every game. What is needed is a publically available program which takes mov into account.

Tango,

Thanks for linking to that article. You owe me two minutes of my life back.

MGL/9,

Totally. Not to mention, the oddsmakers agree. Westwood is usually in the top 5 favorites in every tournament he plays.