AllBuffs | Unofficial fan site for the University of Colorado at Boulder Athletics programs
This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
-
Prime Time. Prime Time. Its a new era for Colorado football. Consider signing up for a club membership! For $20/year, you can get access to all the special features at Allbuffs, including club member only forums, dark mode, avatars and best of all no ads ! But seriously, please sign up so that we can pay the bills. No one earns money here, and we can use your $20 to keep this hellhole running. You can sign up for a club membership by navigating to your account in the upper right and clicking on "Account Upgrades". Make it happen!
AllBuffs.com Intro To APBRmetrics
- Thread starter Goose
- Start date
Another good link is one that jgisland tweeted to me today. I highly recommend everyone checks it out:
http://www.thebasketballdistribution.com/
Also, if you're on twitter, TZI and JG had a good conversation today discussing the team (and TZI even got Ryan Thoburn involved in another one as well) that's worth your time. I would link to it, but my work recently blocked Twitter (despite the fact that I do in fact use it for work, so hopefully this will be fixed soon) so I can't find the link. Their handles on twitter:
jgisland
TZiskBuff
http://www.thebasketballdistribution.com/
Also, if you're on twitter, TZI and JG had a good conversation today discussing the team (and TZI even got Ryan Thoburn involved in another one as well) that's worth your time. I would link to it, but my work recently blocked Twitter (despite the fact that I do in fact use it for work, so hopefully this will be fixed soon) so I can't find the link. Their handles on twitter:
jgisland
TZiskBuff
jgisland
Club Member
I ran across another site last night that the guy from basketball distribution seems to have some involvement in. It is geared toward the NBA but there is a post yesterday with a great glossary/explanation of nearly every advanced stat and method out there. If you are interested in this stuff, read this.
http://thecity2.com/2011/12/01/the-citys-advanced-statistical-primer/
http://thecity2.com/2011/12/01/the-citys-advanced-statistical-primer/
jgisland
Club Member
I will be the first to admit stats can be very misleading, you can backup completely insane arguments at times with some random stats taken out of context or analyzed in a vacuum. You always need to look at any stat and ask yourself, "does this pass the eye test?"
This article by Drew Cannon in Basketball Prospectus does a great job of analyzing stats taken out of context and passing the eye test. I highly recommend the read.
This article by Drew Cannon in Basketball Prospectus does a great job of analyzing stats taken out of context and passing the eye test. I highly recommend the read.
That was a great read.
On a similar note, I've always thought the "pace" number was misleading. Or maybe not so much misleading as it is used incorrectly. It makes sense as a modifier to total stats (points, rebounds, assists, etc.) and allows us to compare Wisconsin to Kentucky on a more apples-to-apples basis.
But what it doesn't really do is tell you whether a team plays a certain style. To get there, we'd really need a "relative pace" stat that told us how much a team plays above or below the standard pace of the opponents it has played in those actual games. "Relative pace" could, for example, show us that a certain team from the BiG is actually pushing pace despite it being below the national pace average due to the opponents it plays. If a team that fit that profile and you were playing them in the NCAAt, the pace stat might tell you that they like to play slow. But relative pace would give you the more accurate assessment that if you try to speed things up you're actually playing right into their hands.
jgisland
Club Member
Not to continually pimp kenpom.com, but he mentioned massey ratings today in his blog. I had never personally heard of the ratings site, I don't pretend to understand this comparative college basketball ratings page on their site, but it certainly intrigues me.
Does anybody have experience with the Massey site? I am going to do some investigating into those ratings on the site, but it looks like there are about 20 different ratings on the site.
Anybody who follows kenpom.com will find his blog post interesting today, he has been taking a bit of heat for his ratings as he still has Wisconsin so high (2nd) after their past few bad losses.
Does anybody have experience with the Massey site? I am going to do some investigating into those ratings on the site, but it looks like there are about 20 different ratings on the site.
Anybody who follows kenpom.com will find his blog post interesting today, he has been taking a bit of heat for his ratings as he still has Wisconsin so high (2nd) after their past few bad losses.
Nor Cal Buff
Well-Known Member
Trending is pretty overrated unless there is a significant change to the team like getting a player back from injury or losing a player to injury. The last game is more relevant than the first game, but I think you'd be surprised by how small the change in relevance actually is.
Nor Cal Buff
Well-Known Member
That is a great blog post. Especially the part about "deserving" to be #2. I wish we had someone/thing like Kenpom for football, too (Sagarin predictor is OK for football, but nowhere near the same level of in-depth analysis).
jgisland
Club Member
Anybody who follows this thread jump on twitter right now and follow @sloansportsconf and follow #ssachat. Sports analytic's chat. Dean Oliver, Daryl Morey, Rich Bucher and many others participating.
Done.
jgisland
Club Member
Found this interesting:
Q) @SloanSportsConf how large a role does analytics play in NCAA sports? HS statistics readily avail in today's world... #SSAChat
A) @ESPNStatsInfo Hard to use analytics for HS plyers but it has started, see krossover. But analytics for tactics - yes
FYI - @espnstatsinfo is being manned by Dean Oliver today
Q) @SloanSportsConf how large a role does analytics play in NCAA sports? HS statistics readily avail in today's world... #SSAChat
A) @ESPNStatsInfo Hard to use analytics for HS plyers but it has started, see krossover. But analytics for tactics - yes
FYI - @espnstatsinfo is being manned by Dean Oliver today
jgisland
Club Member
It was a big weekend in the world of sports stat geeks, the MIT Sloans Sports Conference took pace in Boston. ESPN's True Hoop Blog had nice coverage that is worth checking out. On Twitter #ssac was used as the hashtag to post about the conference, lost of good twitter conversations went on.
Wayne Winston, (former Dallas Mavs stat guy and University of Indiana Professor) has as much of a take down piece as you will get in the stats world. Going after ESPN's John Hollinger's PER.
Wayne Winston, (former Dallas Mavs stat guy and University of Indiana Professor) has as much of a take down piece as you will get in the stats world. Going after ESPN's John Hollinger's PER.
jgisland
Club Member
TeamRankings.com is having a Stat Geek Idol. A "stats blog off" if you will. To a surprise of probably nobody I find this stuff fascinating.
But there was a article called Lehigh, Duke, And The Equalizing Power Of Free Throws.
This part I found really interesting.
Seeing as there are typically 34 fouls called per 132 possessions – or roughly 25.7 percent of possessions (as not every foul ends a possession)
According to this analysis, just over every three extra foul calls are connected to a one-point swing in favor of the underdog. As minimal as it may sound, a lone point is equivalent to approximately 3.5% of win probability according to Pythagorean expectations (such as those used on KenPom.com).
The underdog hasn’t seen such happy results in all the high-foul games this year. Ohio State, Kansas State and Louisville all covered the spread (albeit by less than two points) in games with more than 40 fouls. But 16-seed UNC-Asheville nearly beat Syracuse in a 37-foul game and 15-seed Norfolk State pulled off its own improbable upset in a 39-foul game. 11-seed Colorado won their second-round matchup in a 40-foul affair as well.
But there was a article called Lehigh, Duke, And The Equalizing Power Of Free Throws.
This part I found really interesting.
Seeing as there are typically 34 fouls called per 132 possessions – or roughly 25.7 percent of possessions (as not every foul ends a possession)
According to this analysis, just over every three extra foul calls are connected to a one-point swing in favor of the underdog. As minimal as it may sound, a lone point is equivalent to approximately 3.5% of win probability according to Pythagorean expectations (such as those used on KenPom.com).
The underdog hasn’t seen such happy results in all the high-foul games this year. Ohio State, Kansas State and Louisville all covered the spread (albeit by less than two points) in games with more than 40 fouls. But 16-seed UNC-Asheville nearly beat Syracuse in a 37-foul game and 15-seed Norfolk State pulled off its own improbable upset in a 39-foul game. 11-seed Colorado won their second-round matchup in a 40-foul affair as well.
Nice find... Really neat actually. ****ing Refs. YMSSR
jgisland
Club Member
I imported a bunch of foul data into excel the other day and actually looked at it today.
So what CU games were affected by the refs/high foul counts this year?
Figuring the average is 34 fouls a game, I limited the criteria to anything over 37 fouls. I then subtracted 1 point from the underdog's final score for every 3 fouls over 34. I then looked if the outcome would have changed (none did) but 4 games changed against the Vegas Line/Ken Pom projection. All of those games were in the opponents favor, (meaning they were closer than they should have been and helped the underdog cover). The only game that was actually in CU's favor was the UNLV game, it didn't actually affect the outcome but it definitely helped, I included it in the table.
Here is a table of the outcome: the Vegas Line/Ken Pom Projections is from CU's perspective.
| Opponent | CU Score | OPP Score | Total Fouls | Vegas Line | Ken Pom Projection | Underdog PTS Gained By Fouls |
| UGA | 70 | 68 | 40 | -3 | -1 | 2.00 |
| FRES | 71 | 64 | 38 | -9 | -6 | 1.33 |
| ORE | 72 | 71 | 43 | -6.5 | -5 | 3.00 |
| UTAH | 53 | 41 | 41 | -13 | -16 | 2.33 |
| UNLV | 68 | 64 | 40 | 5.5 | 4 | 2 |
Interesting how the refs almost always favored the other team. Assholes. However, this finding will validate DBT.
Statman
New Member
Yet - Winston (& Sagarin) get the axe from Mark Cuban a few years ago - and Hollinger recently becomes the Vice President of Basketball Operations for the Memphis Grizzlies. Go figure.
This appears to probably be the most appropriate thread for me to make my first "AllBuffs" post....
I was directed over here by an admin after he saw some of my posts elsewhere (and perused my site) - he said there are a handful of statheads (hoopsnerds?) here - gotta love it.
I'm more than familiar to pretty much all the stuff in this thread - as well as a guy that contibutes from time to time at the APBRmetrics board. I like how some are trying to apply some of the more advanced metrics to your Buffs college basketball experience.
To that, I have a site that many of you might find of some interest. I am attempting to do player "ratings" (college and NBA) that a more general basketball fan can look at without getting too confused. My work incorporates many of the theories you see in the advances basketball metrics - with the goal to get everything down to one simple rating that incapsulates that player's performance - everything considered (SoS, pace, playing time in relation to team quality, etc.).
Think of my HoopsNerd rating (and other offshoots from it) as PER (the roots of the metric are linear weights) - IF PER adjusted more effectively for pace - and adjusted AT ALL for Strength of Schedule and individual playing time in relation to overall team quality.
The important thing, for a real more stats oriented college (or NBA) basketball fan - I use my metric to rank every player in D1 basketball (or NBA). I've been doing this for many years for fun actually, but recently have decided to step up and be more serious about it - despite seriously still lacking extra time in my life - and get my site going.
So, please check my stuff out - I'll be doing more blog entries there as the season progresses - and you'll probably see a ton of stuff come tourney time. I have eventual big plans for my for now very green site (neither I or my wife reallyknow much of anything about website stuff). The college basketball player ratings/rankings will be updated every Saturday (or, in worst case scenerio - very early Sunday mornings). The link to the complete college player rankings are at the top of the site (2012-2013CollegePDFs).
And, I apologize in advance, I must admit I'm not a Colorado fan. Don't hate you guys at all, mind you (you ain't Sun Devils), but I am an alumn to a rival you guys are very much hoping to beat very soon.....
Dan
www.hoopsnerd.com
Last edited:
Welcome, statman/Dan. I love the idea for your project. Looking forward to seeing a PER laid out for college hoops in the way you describe.
Now we just got to get you over your self-destructive allegiance to an inferior basketball program. :wink2:
Now we just got to get you over your self-destructive allegiance to an inferior basketball program. :wink2:
soo, what does your system say about Dre and the Mayor? :smile2:
jgisland
Club Member
Good stuff Dan and welcome. I have seen your site before and really like how you have individual comparative ratings for NCAA players, not many people doing that kind of work for the NCAA.
Top 100 in NCAA for both. Dre at 30 and the Mayor at 99.
Top 100 in NCAA for both. Dre at 30 and the Mayor at 99.
Statman
New Member
http://www.hoopsnerd.com/2012-13CollegePDFs.html
100 is pretty much an average D1 player. Roberson was the #1 ranked P12 player and the #30th ranked player nationally with a 160 Hoops Nerd Rating thru last Friday's games. Dinwiddie was ranked 8th in conference and #99 nationally with a 146 rating.
The Buffs were one of 18 teams to have at least two players ranked in the top 100 players nationally. Arizona was the only other P12 team to have two (Hill & Nick Johnson).
Wow. Thanks man. That's some posrep for you. Also, how bad must our bench be to be in this rarified air as far as top end talent but have the losses we have?
Also gives me hope against Zona. Thanks!
jgisland
Club Member
How much do your projections deviate on a week to week basis? I see Dinwiddie's rank went down 11 spots from 88 to 99 this week. In that week Dinwiddie had the "perfect game" of 24 pts 6-6 FG, 4-4 from 3 and 8-8 from the line so I find that surprising. Is there anything in particular you can attribute that drop to?
Statman
New Member
You beat me there.....
I believe I am one of a grand total of two guys that even attempts to rank all D1 players and post complete results....
The other: http://www.valueaddbasketball.com/
His dataset has some definite flaws (players on wrong team, etc) - and some of his results end up quite skewed, especially with lower efficiency volume shooters. I've taken some flack from people who think Jahii Carson is some elite level PG - when my ratings have him ranked 490th in the nation (WELL above average D1 player, mind you), and moving up slowly (almost passed their 8 foot center as the 2nd best player at ASU this last update). The site above currently has him ranked 895th in the nation and 4th on ASU (and much closer to being the 7th best player on ASU than 2nd).
Shabazz Mohammad, another guy people seemed determined to still call elite - ranked #341st nationally in my metric, #505 in the site above.
I'm obviously hoping that a major media might catch on and realize that people may find it interesting to see how their favorite player's ranking may have changed from day to day (updating player rankings daily, imo, would be awesome) - just like (although maybe a lesser extent) somehow millions of people seem obsessed with early season "bracketology" - projections that are honestly, kinda, complete guesses. Yes, I'm bitter and very jealous of Lunardi - although it's probably his winning personality and stunning good looks....
Pomeroy kinda ranks college players - he has kPOY - but he'll only post top 10. I believe this is because he knows he's take some flack for his results if he posted more - he had THREE Florida guys ranked in his top 10 nationally much of the season. When they started Basketball Prospectus about 5 years ago - Pomeroy and I talked on the phone for a good hour about player ratings (I was the ONLY guy anyone knew that even attempted to rank all the players). They later decided to not bring me on as a contributor, since Pomeroy was going to be their college bball guy - and he supposedly had an overall player ranking system in the works. Still waiting for it.
People don't tackle this because you REALLY have to love college basketball to go through all the work - and it's REALLY difficult to get results that pass the eye test most of the time for EVERY player. I doubt anyone will come up with rankings that "look" better than mine anytime soon (ever?) - especially when I get the historical stuff posted in a few months and people can see 15 seasons of results.
Just wait for future college projections, college to NBA projections, and future NBA projections - then things will get real.......
Statman
New Member
Actually - the update was BEFORE that game (thru Friday's results - the game you are talking about was this last Sunday).
His week that saw his rating drop from 149 to 146: bad road loss (Utah) and tough (important) road win (Oregon) : 13 ppg, 0.5 rpg, 3.5 apg, 3.5 topg, 0.5 spg, 0.5 bpg, 52% TS%.
Wasn't one of his best weeks.
The ratings can change quite a bit from week to week - blowing up statistically while your team gets a couple HUGE wins can bump a rating big time (especially if your rating was weakish to begin with) - struggling while your team gets upset twice can be a kinda large fall (Kansas players?). Look at the recent nationally rankings - scroll it - you can see the rating change in the last column, it'll give you some insight on how much the ratings can change.
The biggest upswing that last week (2-1 to 2-8) of a guy that was nationally ranked (top, I dunno, 700) was BJ Young of Arkansas: his rating went up by 8.7 and his ranking went from #232 to #135. Arkansas handled Tennessee and FLORIDA that week - and he was 19 ppg, 3.5 rpg, 4 apg, 0.5 topg, 66% TS%. Great stats with a nice team rating bump (upsetting Florida) equals big player rating bump.
Biggest DROP of a highlyish ranked guy - Damion Lee of Drexel. HUGE rating drop of 12.1 and a ranking drop from #140 to #298. Road loss to a mediocre team, home loss to a BAD team (2-20 ODUM) - he was 3.5 ppg, 2.5 rpg, 1.5 apg, 3 topg, 4 pfpg, 0 steals or blocks, 19% TS%.
