Wow – eight for eight for the favorites. That in itself has to be unexpected. There were a few close runs, including a very exciting attempt at history by #16 seed UNC-Asheville – but in the end, the chalk has moved on. And that means that we have 116 perfect brackets so far. Just amazing!!!! Standings look to be tight at the end of the night. Maybe VCU, South Dakota State, New Mexico State, or Colorado can help provide some excitement for those of us who crave upsets. (sorry – I can’t consider Connecticut an upset – even Vegas is listing the 9 seed as the favorite in this game).
2 responses to “No upsets so far”
“That in itself has to be unexpected” The 8 for 8 favorites winning beginning is not really unexpected. If you thought each game was a 50-50 chance then this particular pattern is 1 in 2^8 or 1 in 256. So, yes, that is a bit weird. But in that case, there really weren’t favorites. So, the chance of these 8 teams winning depends on what you think each game’s probability is. My BT model is in the space of sensible (yeah, not perfect. Sometimes I think it separates the teams too much, and other times too little. But it is generally in the ballpark (or in the court I guess)). When I multiply the 8 probabilities for these games I get a chance of 18% that the favorites would win all 8 of the games. (the 1/8th root of .18 is .81. (now, isn’t that a cute piece of numerology there) So, on average, the model thinks the higher seeds in these first 8 games have a .8 chance to win) So, 8 for 8 is not really that unexpected. Certainly it is in the space of the sort of thing that happens regularly. The BT model is says there is only a .0002ish chance of all the 32 favorites winning in the second round. Now, that would be unexpected and would suggest wildly unevenly matched games (with the averge games being like a 90% certainty) or some sort of lack of iid (e.g. collusion on some grand scheme). But such has not happened and all is playing out nicely with respect to the simple model. Any other expectations your want calculated? I am enjoying testing the BT stuff in these many ways. -Bill
Tom–You are so right. Very tight grouping in this pool. Over in the 2.95 million entry Yahoo pool my 14 for 16 has me in the upper 3rd percentile. In your pool I am in the upper 24th percentile. Can you imagine–14 for 16 and only upper quartile? This is a group of prognosticators that are seriously out to win. I guess, collectively, this is a group of some of the world’s best large-scale risk managers. (Or, at least, those who love (or at least tolerate) us.). Certainly seems like we have made much more conservative (and successful) picks than the large Yahoo pool. -Bill