tag:blogger.com,1999:blog-22972768.post6921814669633978367..comments2023-09-01T06:51:04.030-04:00Comments on Basketball Predictions: W-13 BP68Jeffhttp://www.blogger.com/profile/12767101715737129123noreply@blogger.comBlogger49125tag:blogger.com,1999:blog-22972768.post-79860449589212275052011-12-14T15:16:08.690-05:002011-12-14T15:16:08.690-05:00Right. But over time it will even out. If every t...Right. But over time it will even out. If every team played 100 games, the result of any one game is going to be but a rounding error. Even once these teams play 15-20 opponents, the numbers get pretty rock solid.<br /><br />That assumes, of course, that the teams are "connected". This is one of many flaws with the BCS model in college football (besides the flaws you already know about, and besides the fact that the computer polls they use don't take into account the scores of games) - the fact that the top conferences don't even play each other. Because college football rewards you for putting together the easiest schedule possible, the top teams in the top conferences basically never play. There will be two or three games per year between two teams from the top half of different BCS conferences. It makes it very difficult for the computers to accurately measure the relative strengths of the conferences. Is the SEC the best? The Big 12? The Big Ten? It's impossible to tell because they don't play each other.<br /><br />We don't have that issue in college basketball. The top conferences play each other often enough that the computers can fairly accurately measure how good teams are and how good their resumes are.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-85881932732536278472011-12-14T15:11:52.219-05:002011-12-14T15:11:52.219-05:00I think I understand, basically those teams are li...I think I understand, basically those teams are likely a little over ranked in his system than they should be due to those games taking place in such a small sample size...correct?Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-22335109866516346522011-12-14T15:07:13.888-05:002011-12-14T15:07:13.888-05:00I think you misunderstood Pomeroy's point. Wh...I think you misunderstood Pomeroy's point. What he's saying is what I'm saying, which is that early in the season you have sample size issues. If one team has a horrible game where none of their shots fall, they'll be under-ranked. If a team has a fluke game where everything's clicking, they'll be over-ranked.<br /><br />As the sample size gets larger, each individual performance starts to wash out.<br /><br />With Pomeroy and Sagarin, the computer ratings don't start being very accurate until mid-January or so, so anything at this point has to be taken with a grain of salt. Every time I use the stats I make that clear, and that's why you'll notice that there isn't a whole lot of overlap between my bracket and between any of the computer ratings.<br /><br />Sagarin doesn't reveal his exact formula, but it's based off the classic Elo rating system that you'll see all over the place for an array of different sports. It was originally invented for chess players (hence the name "ELO_CHESS". You can read about the basic structure at wikipedia:<br /><br />http://en.wikipedia.org/wiki/Elo_rating_systemJeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-61011057902897761712011-12-14T14:58:50.097-05:002011-12-14T14:58:50.097-05:00Interesting stuff Jeff.
How does Sagarin come up ...Interesting stuff Jeff.<br /><br />How does Sagarin come up with the ELO CHESS? I didn't see a breakdown of what exactly that means on his site. <br /><br />One point I forgot to make yesterday in my argument against Purdue is that Ken Pomeroy said that due to the huge win against N. Illinois, Purdue will be forever overrated in his system. He went on to say Wisconsin and Marquette are in the same boat as Purdue. <br /><br />Just a thought.Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-51823217276610998502011-12-14T12:24:40.302-05:002011-12-14T12:24:40.302-05:00Dude- I understand stats and computer rankings. I ...Dude- I understand stats and computer rankings. I "get it". I'm trying to clue you in with a little relevant info... at some point common sense has to take hold<br /><br />ELO chess teams ahead of us...<br />Murry State = 5<br />Ohio = 10<br />Mississipp = 15<br />Southern Ms = 24<br />Northern Iowa = 22<br /><br />This is not an exhaustive list, but clearly the computer ranking don't mean much at this point. I can't speak for all those teams, but the other MS schools project better than us?? It's completely nonsensical.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-29393403709793528072011-12-14T10:32:57.163-05:002011-12-14T10:32:57.163-05:00You should go back and read what the others on her...You should go back and read what the others on here are saying, because we're answering all of your questions. The fact that you guys beat teams seeded ahead of you does not matter. Nearly every team in the nation will end up beating at least one team that ends up being seeded ahead of them in March.<br /><br />Here's one more stat for you - the Sagarin ELO_CHESS. The ELO_CHESS is the most accurate metric of resume strength, which is what teams are seeded on in March. If you want you can go to the bottom left of the main blog page and click on any of the "How well did the computers predict the field" posts to see just why the ELO_CHESS is the closest objective metric to the eventual Selection Committee. This metric does not take into account the scores of games, it doesn't take into account if a team was "hot" or "cold" shooting, or whether a November tournament is "weird" or not.<br /><br />Anyway, what is MSU's ELO_CHESS right now? 51st. I can tell you right now that only three or four teams with an ELO_CHESS outside the Top 50 on Selection Sunday will end up earning at-large bids.<br /><br />So MSU's resume, right now, is <i>right</i> on the bubble. As DMoore said, they're "bubblicious". They'd be one of the last teams in or one of the first teams out. In other words, the only way I can project MSU into the field is if I think that they've either performed exactly how they're going to perform the rest of the year, or if I think they've <i>under</i>performed. And I don't think either is true. I think they've slightly overperformed. And so I have them narrowly outside the bracket.<br /><br /><br />By the way, remember what I said about projections? Texas A&M is currently 127th in the ELO_CHESS. No team with an ELO_CHESS over 100 has ever or will ever get considered on Selection Sunday (unless they expand the field to 96 teams or more). But everybody has Texas A&M in their bracket, including me. Why? Projections. We know A&M is better than that, and they've just played a soft schedule so far.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-51591063081100398322011-12-14T09:52:00.909-05:002011-12-14T09:52:00.909-05:00Rob Anon here -
I'm not suggesting that we w...Rob Anon here - <br /><br />I'm not suggesting that we win out over the rest of the season, just that having us out of the tourny at this time can not be based upon anything logical. That's all. I don't even object to a low to middling seed. Again, the only reason this has gone back and forth are the dichotomous inferences that we "got lucky against WV", experienced "holiday tourny mojo", yet suffered a single "bad loss" that does however exemplify us. There are suggestions that we aren't beating teams bad enough, yet we are "clicking" now. The teams we have beaten aren't all that good, but with lesser resumes they are "in", we are "out". <br /><br />there are a lot of teams out there, so I do not expect anyone here to have internal knowledge of our team.. that's okay. I enjoy the banter, but gotta call out silly logic when I see it.<br /> BTW, if Kenpom means so much, then just use his rankings rather than your own. <br /> To bolster your inside nformation on the dawgs...<br /><br />I'm not suggesting that we are unbeatable..however to DMoore - yes Dee is important, but we have a very good back up available when dee isn't on his game... Deville Smith. He sat out the Akron in a hospital due to an illness.<br /><br />Dee is huge - but Arnett Moultrie is our inside beast. FYI, no double digit win last night, but note that Moultrie sat out last night (for the 3rd time against "lesser teams") due to tendonitus.<br /><br />Clearly, the coach doesn't realize the impact of these non-blowouts.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-87860180859558340682011-12-13T22:27:17.838-05:002011-12-13T22:27:17.838-05:00Dee Bost is a stud. Can't disagree with that. ...Dee Bost is a stud. Can't disagree with that. The team rises and falls with his play.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-46946509556673942242011-12-13T22:18:30.500-05:002011-12-13T22:18:30.500-05:00Anonymous MSU fan-
OK. Looking at MSU's recor...Anonymous MSU fan-<br />OK. Looking at MSU's record so far, I see three good wins (West Virginia [home, KenPom 32], Arizona [neutral, 41] & Texas A&M [neutral, 56]) and one bad loss (Akron [home, 110]). Also, MSU's KenPom ranking is 50.<br /><br />That's not bad, but by end of the season standards it is bubblicious. With those scores, beating Arizona and Texas A&M are wins over other teams who also appear on the bubble. The WV win is nice, but it's at home, and it's balanced out by a bad home loss. And, the team has zero road wins (hasn't even played a true road game).<br /><br />So, there is nothing special in their record. Whether they are on the good side or bad side of the bubble would have more to do with the comparative records of the other bubble teams.<br /><br />Using the eye test, I would put them on the good side of the bubble, mostly because I think Dee Bost has been having a really nice season so far. I think it should be pointed out that when he has a bad game, MSU can lose to inferior teams (Akron).DMoorenoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-76753721195433175052011-12-13T21:39:47.088-05:002011-12-13T21:39:47.088-05:00Indiana just seemed to get all the 50/50 calls. K...Indiana just seemed to get all the 50/50 calls. Kentucky was the team slicing through the lane and drawing contact - they should have had many more free attempts than Indiana.<br /><br />It wasn't particularly egregious. There is almost always some home cooking from the refs at all arenas - it's the dominant cause of "home court advantage". That's why Vegas gives home teams an extra 4 points in the spread.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-23979638317456763402011-12-13T21:33:58.515-05:002011-12-13T21:33:58.515-05:00Actually I'm a Michigan St. fan, that's wh...Actually I'm a Michigan St. fan, that's why I haven't really mentioned them....Ha. I grew up in Indiana though and couldn't stand either IU or Purdue. These days I just find myself rooting for the Big Ten in general because I watch them all the time.<br /><br />One note on the UK/IU game though. Didn't both teams shoot 17 FT's? Where was the home cooking?Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-4290906774523052592011-12-13T19:13:04.170-05:002011-12-13T19:13:04.170-05:00I'm not sure what your point is. At the end o...I'm not sure what your point is. At the end of the season every win will count be counted. But you seem to be saying "So far we've beaten every quality team we've played, so treat us as if the whole rest of the season we'll beat every team we play except for Kentucky and Florida." I just don't see why it makes me dumb for thinking that MSU won't play as well the rest of the year as they've played so far (and since both Sagarin and Pomeroy rate them around 50-60th, it can be argued that they actually need to <i>improve</i> to make the Tournament).<br /><br />This is particularly the case since wins late in the year matter more than wins early in the year. If Mississippi State goes 7-9 in SEC play they won't earn an at-large bid - the fact that they beat Arizona on a neutral site in November won't matter.<br /><br />This week, MSU plays Florida Atlantic at home, and a road game at Detroit. If they can win both games by double digits I'll be impressed and there's a decent chance I'll move them into the bracket. But I'll tell you right now - the odds are strongly against them winning both games by double digits. Just winning a game at a Detroit team that just got Eli Holman back is going to be a difficult proposition.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-74336527120692912712011-12-13T18:58:47.866-05:002011-12-13T18:58:47.866-05:00You have your site and your right to post anything...You have your site and your right to post anything. I simply asked you to justify your selection as it relates specifically to MSU and the teams we have BEATEN. You really can't, so rather than say we'll fold, the only thing mindblowingly stupid is diminishing legitimate wins on a neutral site and claiming one team lost because they were ice cold and we are "clicking". I still don't think you watched the game or you wouldn't suggest it was just a matter of WV going cold. It was a good game and we won. WV 3 seed, MSU out?. Justified or not, someone keeps voting A&M in the top polls - you have them in.. and we beat them also. Enough! A legitimate ranking system has to have some rational and all you can go by is what has actually happened over the course of a 3rd of a season...this isn't one game where we "were clicking". Given what appears to be the relative strenght (over recent years), I would expect that a 9-7 record might in fact get us into the tournament. You don't think 22-9 would be a shoe in given that we would sure to have a decent rpi one would think, especially with Baylor coming up soon.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-45623004089474879432011-12-13T18:15:57.556-05:002011-12-13T18:15:57.556-05:00Haha, yes, I'm trying to be clear at the top o...Haha, yes, I'm trying to be clear at the top of each comment who I'm responding to, but it's not always clear.<br /><br />Anyway, I'm not clear what my "oversight" is on Mississippi State. Right now Pomeroy projects them to finish 21-10 with a 9-7 SEC record, and RPI Forecast (which uses the Sagarin numbers) projects a 22-9 finish with a 9-7 SEC record, so it's not like I have to be mindblowingly stupid to think Mississippi State isn't a shoo-in for the NCAA Tournament.<br /><br />Like I said, I just think the sample size is too small for Mississippi State. They have three wins over quality teams, and all three have a caveat. Two were at a November tournament which, as I said, are notoriously unreliable measure of future performance. And the West Virginia win came in a game when WVU couldn't hit a shot. It was a game where everything was clicking for MSU and everything was coming up garbage for WVU.<br /><br /><br />Here's the thing with small sample sizes. The only way to project the future is to try to look at the games and figure out what each game really means. We know that teams have days where things are clicking and days where they play badly - this is true of every team. If a team gets a big win, was it because everything was clicking, was it because their opponent was ice cold, or was it just a solid victory? In my opinion, MSU is a team that's been clicking. I think they've over-performed so far, and will return back to earth as the season goes along.<br /><br />Again, that's just my opinion. Maybe I'm wrong. Wouldn't be the first time.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-29830997201394597382011-12-13T17:32:31.858-05:002011-12-13T17:32:31.858-05:00If you are talking to me - I'm Rob the bulldog...If you are talking to me - I'm Rob the bulldog fan - This is all hocus pocus this early anyway, but you might put relevant weight on the outcome of games played against decent teams. Look, we beat Troy by nearly 40 a few days ago, so I think you are justifying your ommission. another thing - sure it's reasonable to exclude a game one way or another when judging a single team - meaning one team played exceedingly poor (ice cold,injuries, got screwed, and so forth). However, State (MSU) has played 3 top 32 teams (according to you) and has won all three. They indeed have a poor loss against an Akron team (I haven't watched them since, but shocked because they looked good against us), but we haven't been the same since and Deville was out that game also. The point is, even with the "bad loss", you'd think we are a shoe in for the the top 65...... especially since you qualify in the top 32 those same 3 teams that are 21-6 with half those losses against us. We could lose tonight and the point is moot I suppose, but to this point, you'd have to argue we are in. If you'd just acknowledge your oversight, I'd be more forgiving than trying to justify what doesn't make sense.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-72812102084483823402011-12-13T16:55:44.494-05:002011-12-13T16:55:44.494-05:00Anonymous, how much weight should be put on what&#...Anonymous, how much weight should be put on what's been done so far, and how much weight should be put on projections of the future? 60/40? 70/30?Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-23710925862234164612011-12-13T16:54:52.443-05:002011-12-13T16:54:52.443-05:00Charles, I'm starting to get the hint that you...Charles, I'm starting to get the hint that you're a Hoosiers fan, and that's clouding your opinion on Purdue. Pomeroy rates Purdue as the 14th best team in the country, thinks Butler is awful, and still only has Purdue as a 9 point favorite. So saying Purdue is only favored by ten because they "aren't very good" isn't true.<br /><br />And by the way, while I did pick Indiana to beat Kentucky (and put my money where my mouth was), if anything Indiana is overrated right now because of how high profile that one win was. I picked Indiana because I expected a very young Kentucky team to panic and loose their cool in their first true road game, and because I expected Indiana to be at their best. Both of those were the case, yet even with that and some home cooking from the refs it still took a miracle three-pointer for Indiana to win.<br /><br />Indiana had absolutely everything clicking in that game - it's unreasonable to expect the rest of the season to go that well. Or do you plan on Indiana hitting 60% of their threes this season?<br /><br /><br />Here's a way to think about it. Home court in college basketball is assumed to be worth four points. Purdue is playing a "virtual road game" against Butler, since the game is in Indianapolis, so we can treat that as being worth two points. If Purdue is favored by 10 in this game, it would mean they'd be favored by 16 if Butler came to play them at West Lafayette.<br /><br />Indiana already played Butler at home... and won by exactly 16.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-63886913994268299862011-12-13T16:48:34.713-05:002011-12-13T16:48:34.713-05:00Projecting what the actual field will look like in...Projecting what the actual field will look like in March is virtually impossible right now unless you go through every single team and predict win/loss for each game every week. <br /><br />Projecting teams on what they have done seems to make a lot more sense. Your example of Santa Clara over NC isn't really a good one because you have to be both subjective and objective when doing this. And I really hope no one would take the time to project the field after 1 game. Impossible to give it validity.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-87519954411501232972011-12-13T16:27:44.899-05:002011-12-13T16:27:44.899-05:00The reason Purdue is only around a 10 point favori...The reason Purdue is only around a 10 point favorite against Butler is because they just arent very good this year. Butler is awful.Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-47825494935722347972011-12-13T16:26:16.835-05:002011-12-13T16:26:16.835-05:00Again, it depends on whether or not Hummel is incl...Again, it depends on whether or not Hummel is included in the bunch. Ryne Smith does only one thing, and when teams get a hand in his face he is off. I like Jackson when he doesnt make bone headed mistakes. <br /><br />How about Indiana then? They have beaten every team by double digits other than UK. If they wouldnt have gotten away from what got them there they would have won by 10 plus. They arent getting nearly the respect they deserve by most people. Likely because of how down they have been the last few years.Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-81570762745532239112011-12-13T16:12:06.809-05:002011-12-13T16:12:06.809-05:00I'm not sure I'd say Purdue "should&q...I'm not sure I'd say Purdue "should" win by double digits. Pomeroy and Sagarin, which both rate them as a Top 25 team and Butler as well outside the Top 100 have Purdue as a 9 and 13 point favorite, respectively. I haven't seen a Vegas line for the game yet, but I'd expect it to be right around 10.<br /><br />A double-digit win would be a good result for Purdue, but I wouldn't say it's "expected".Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-54747667786706579522011-12-13T16:08:58.143-05:002011-12-13T16:08:58.143-05:00Really, Charles? Which three Big Ten backcourts wo...Really, Charles? Which three Big Ten backcourts would you take over Purdue? The only one I think you can make a good argument for is Wisconsin.<br /><br />And the point about cupcakes isn't that you need to blow every single one out. The point is that it's all part of the sample. Each game matters. Teams will have good days and bad days. Overall, Purdue has been better than MSU so far - neither Pomeroy or Sagarin have them even close. And like I said, I feel like Purdue has underperformed while MSU has overperformed. I think MSU got lucky that West Virginia just happened to have a horrid shooting day that game. Over the long run that stuff tends to even out.<br /><br /><br />And anonymous, if you aren't familiar with the concept of "guts vs stomps", let me recommend you some <a href="http://www.footballoutsiders.com/stat-analysis/2005/fo-fox-guts-and-stomps" rel="nofollow">good reading</a> to start with.<br /><br />Despite what our intuition tells us, stomping all over bad teams has a better correlation with future success than whether a team wins or loses close games against quality opponents. It's true across all major sports.Jeffhttps://www.blogger.com/profile/12767101715737129123noreply@blogger.comtag:blogger.com,1999:blog-22972768.post-57271542029571146722011-12-13T16:04:25.461-05:002011-12-13T16:04:25.461-05:00I'm interested to see how Purdue plays this we...I'm interested to see how Purdue plays this weekend against Butler. This is a game they SHOULD win by double digits.Charlesnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-47747646181257889532011-12-13T16:01:02.993-05:002011-12-13T16:01:02.993-05:00I said that Akron was a bad loss, dude.... but I d...I said that Akron was a bad loss, dude.... but I do expect a little logic in your reasoning. <br /><br />first, you lop off our best 3 games (all wins) and declare that "wierd things happen at holiday tournys" and suggest that WV SHOULD have beaten us. Did you even watch the game or just look at the stat line? We couldn't have played any worse against Akron, so by that reason, let's toss that game out also. <br /><br />I've never heard anyone reflect on margin of victory in cupcake games. That's bogus, because we haven't been challenged in the least since the Akron game (game 2). Our scrap team guys (the guys you ONLY see in blowouts) have had more PT than I can ever recall, but who knew that guys like you deduct points because they can't score worth a damn. Again - you see the 3 teams we beat as worthy of the top half of the bracket, yet we are out? You can't justify that by anything logical...it's inconsistent. If I had time, I'm certain I could apply the same logic you've used and exclude half of your bracket. Seriously.. "wierd things happen"?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-22972768.post-68934096092261435382011-12-13T16:00:33.194-05:002011-12-13T16:00:33.194-05:00I'm in Big 10 country and have watched Purdue ...I'm in Big 10 country and have watched Purdue a lot. I can assure you they do not have the best backcourt in the conference, and zero inside game whatsoever. If you throw Hummel in as part of the backcourt then maybe they are in the top 4 or so. <br /><br />I didnt mean Temple and Iona were cupcakes. I meant High Point and Western Carolina. It was in response to your claim Miss St has some unimpressive wins.Charlesnoreply@blogger.com