Tomorrow the Pac 12 Awards will be announced. As an Arizona alum and an unabashed fan, I and my fellow Wildcat fans are hoping that our senior point guard, TJ McConnell, wins Pac 12 POY. Fingers crossed.
But, I’m a basketball analytics freak who believes that, after everything is properly accounted for, the proof is in the pudding. Despite my incredible bias and knowledge that he has been playing exceptionally in conference, does my main man TJ really DESERVE the POY honor? Well, I decided to take a few hours of my Sunday afternoon to find out.
Thanks to sports-reference, it is very easy now to pull college player data that is accumulated only in conference games. I assume since we are trying to put together the best possible list of the statistically most deserving Pac 12 players – unlike my weekly national rankings it is probably best to use only Pac 12 games when crunching the numbers.
Now, this is different than my normal college player rankings. I love to use WAR in my NBA ratings, so I think I’ll do the same here. A fairly even strength of schedule among the Pac 12 teams (I know, not perfectly even, but when is anything ever perfect) allows me to simplify my calculating of relative team strength as TmPts/OppPts*100 (just Pac 12 outcomes obviously). 100, then, would be a completely average Pac 12 team. I don’t have the patience to run regressions at the team level to get a more “precise” team rating, but since everything with WAR gets tied back to actual team wins – the slight noise of the team ratings smooths out quite a bit. Trust me.
I’m actually approaching these ratings exactly how I would ANY “pro” league with a relatively balanced schedule. My overall desire, just like my NBA ratings, is to have a final WAR, WAR per minute (WAR/40 – it’s actually the player’s WAR projected out as if they played EVERY minute of the Pac 12 season), HnR (the rating I use for the national rankings I do every week – but this will be based solely on Pac 12 results with 100 being Pac 12 average instead of D1 average), and HnI (HnR that ignores missed games, the best at the moment “impact” rating of the player when he’s actually available to play – the one I use for game projections).
The player stats here get “adjusted” to a uniform league norm – this allows for my ability to look at ANY league’s player production as if it were pretty close to current NBA pace and statistical ratios. In this case, assists are a little “rare” compared to the NBA (not as “rare” as most Euro leagues though), while turnovers and especially personal fouls are a bit more prevalent. So, this should help my man TJ a little (as well as Deron Wright), as his best statistical skillset (assist rate and A/TO ratio) stands out more in this season’s Pac 12.
As always with any of my ratings, all player stats are compiled at the team level before they are broken down back to the player level in terms of ratings as each statistical skillset. This properly adjusts then for pace, team defense, maybe having overly generous home scorekeepers, etc.
Finally, WAR should be an interesting result here. In the NBA, teams play 82 games, and over the course of an entire season their team ratings will correlate fairly closely to actual team wins. But in this case, we have just 18 conference games. Utah, based on their team rating in conference of 123, would have been expected to win about 15 games. Circumstances would have it (being “unlucky”, not being clutch,, etc. – just plain losing close games while winning with big blowouts) that Utah won just 13 games. Oregon had a team rating of 103 – a bunch of close wins with some bad blowout losses – yet instead of winning the expected 10 games with that rating, they won 13 just like Utah.
So, while Utah players in general will have a much better HnR & HnI because of a much better team rating, their WAR will be based of the same 13 wins as Oregon. The Oregon players, just as the team, deserve to be given the credit for those wins, even if they are probably based much more on luck than they would be in a much longer schedule (like the NBA) that smooths out the WAR “noise”. Since we are attempting to create a viable all conference team that “looks” at performance like actual Pac 12 voters (all the while adjusting for all the team factors much better than the average voter would) – WAR and it’s tie to team wins and player playing time almost certainly will really help do that. If want to rank the best conference players this season on a points per possession basis, understanding some of the better Pac 12 players are buried behind better players on some teams, we use HnR (where 5 of the top 9 Pac 12 players are Arizona Wildcats, two of the top 6 Pac 12 players are Utes).
But, if we want to reward players based on how much their specific impact probably tied to actual team wins – and while the results may be noisier with just an 18 outcome sample size for each team – we’d look at player WAR. While Oregon did “dominate”, they pulled out 13 wins in 18 games. That means something – and while it’s not predictive of future outcomes, it’s important when rewarding players with 1st or 2nd team Pac 12 honors.
OK, I got all of that out of the way. Here is every Pac 12 player who played even just 1 minute in conference based on WAR – along with all their conference stats, WAR/40, HnR, HnI, and even conference rating percentile rank in various statistical subsets:
There’s a bunch of info there obviously. EVERYTHING I’m about to talk about can be found in that spreadsheet in much greater detail. It can be sorted any way you’d like, to your heart’s desire.
Pac 12 POY
Since the moment you open the spreadsheet it’s staring you right in the face, I’ll just start from the top.
An unexpected leader in WAR – Gary Payton. Oregon State won 8 conference games, despite a conference team rating of just 93. They were a bad offensive team and a good defensive team who played a really slow pace – so nobody statistically is going to look “great”. But, Gary Payton still stands out. Over half the team wins are directly attributed to him, which is amazing. The team rating drops over 13 (from 93 to 79) if we take Gary’s production and minutes out of the team results – EASILY the most important player to his team in the conference.
But, a great player on just an 8 win team won’t win POY. We have two guys who arguably would be more “deserving” of the award – BOTH who played less per game but were THE stars on teams that were loaded with good players – Deron Wright and TJ McConnell. Both had better WAR/40 than Gary Payton (they played quite a few less minutes to get to their respective WAR totals), as well as much better HnR and HnI.
But, much to my chagrin, Deron edges out TJ, in all four metrics I mention. Utah slightly underachieved in terms of wins relative to team rating – BUT Deron Wright was the unquestioned star on a strong Utah team. TJ was the best player on a team full of stars, the perfect floor general to feed the egos of his teammates by feeding them the ball. However, you take Deron from Utah, they drop 10.3 rating points. You take TJ from Arizona, and they drop 6.1 rating points (133 to 126) – and (theoretically) are still the best team in the Pac 12.
Pac 12 POY: Delon Wright of Utah.
POY Runner Up: TJ McConnell of Arizona.
POY Honorable Mention: Gary Payton of Oregon State
ALL that being said, TJ McConnell winning POY would by no means be crazy. His Pac 12 season was exceptional, even if it measures out just a tick below Delon’s by my metrics. He still was the best player on easily the best team – and voters often vote that way. This wouldn’t be voting Kevin Pangos over Kyle Wiltjer as WCC POY – THAT was crazy.
Pac 12 All Conference Team
The Pac 12 awards 1st team All Conference to TEN guys. Weird. I’ll do the same, but I’ll order them into a top 5 and next 5, so it looks better to me.
Deron Wright – Utah
TJ McConnell – Arizona
Gary Payton – Oregon State
Joseph Young – Oregon
Stanley Johnson – Arizona
Joseph Young’s WAR, as I explained before, is much higher than one would expect based on his HnR. But, Oregon tied for 2nd in conference, and he played a TON of minutes as their best player. If I were drafting a team of Pac 12 players to play some important game tomorrow against China (winning the game would relieve the USA of all China debts) – my HnR and HnI (and my personal intuituion) says Rondae Hollis-Jefferson would be the man for this starting spot on the first team. I’m ignoring positions in the fantasy world – just want the five best.
If the Pac 12 played an 82 game season (or even a 41 game season), the high WAR ranking of the Oregon guys would settle down a bit toward their HnR, since their winning % would end up more 55% (based on the team rating) than the 72% it was this 18 game season. On the opposite end, the low WAR of the USC guys would settle up a bit – their team win % would be closer to 33% than the 17% it ended with.
Rondae Hollis Jefferson – Arizona
Chasson Randle – Stanford
Kevon Looney – UCLA
Norman Powell – UCLA
Brandon Taylor – Utah
I diverted from the WAR total here with the last guy – giving Taylor and his slightly lower WAR and WAR/40 the nod over Elgin Cook of Oregon, because Taylor’s HnR and HnI were 20 points higher each. Gotta beat China.
Anthony Brown – Stanford
Josh Hawkinson – Washington State
Elgin Cook – Oregon
Askia Booker – Colorado
Brandon Ashley – Arizona
You’ll notice I ignored the slightly higher overall WAR of Shaquielle McKissic of ASU for the slighly higher WAR/40, and the definitely higher HnR & HnI of Brandon Ashley. The difference in the latter metrics more than makes up for it in my mind – plus, um, I’m an Arizona alum/fan. Need I say more?
Overall – I think the WAR results we get here on their own correlates pretty well in terms of how I see fan and voter perception of the players. Gary Payton is an outlier – his huge minutes with a very good WAR/40 (THE star on a good defensive slow paced overachieving team) getting him the overall WAR edge – but the actual POY will will be one of the top two in HnR. One could fashion the All Pac 12 teams based directly on the WAR rankings – and I don’t think there’d be much (if any) media or fan uproar.
Oh wait – almost forgot….
All Freshman Team
Stanley Johnson – Arizona – FOY
Kevon Looney – UCLA
Jakob Poeltl – Utah
Jordan Bell – Oregon
Dillon Brooks – Oregon
If you are still reading this, and are thinking this Hoops Nerd guy has nothing on so & so when it comes to college basketball &/or Pac 12 analytics – get on twitter and let me know who else may be crazy enough to crunch these numbers. Sometimes I feel like I’m kinda alone here.