Wednesday, May 31, 2017

LeBron vs Jordan, Supporting Casts and Finals Opponents

In the ongoing GOAT conversation between LeBron and MJ, LeBron has frequently been lambasted for changing teams to go to greener pastures with better (and younger) teammates when he couldn't get the job done with his current team. The argument goes that Jordan never left the Bulls (except for that one year he retired to play baseball/serve David Stern's secret suspension for gambling) and won his titles with the players around him. Thing is, those players were pretty good, including two future Hall of Famers. So how did Jordan's counterparts stack up with LeBron's?

I took every season in which either player made the NBA Finals and compared the other starters' average PER on each team, the top two units' average PER (starters and top five bench players), and each star's PER in each season.

SeasonWon TitleStartersTop 2 UnitsPER

SeasonWon TitleStartersTop 2 UnitsPER
*First season on a new team; 2010-11: Miami Heat, 2014-15: Cleveland Cavaliers

The other starting teammates for both LeBron and Jordan were actually fairly comparable on average, but Jordan did have a larger advantage with his bench unit. The gripes with LeBron are accurate though: in both seasons prior to when he left his respective teams in free agency, his teammates were remarkably worse than who he had on his next team. That being said, Jordan never had a situation as bad as LeBron's until his last title in 97-98, so he arguably never had an incentive to jump ship like LeBron did.

Not surprisingly, LeBron's teammates were better in the years he won the title than in the years he didn't, on average. There's a point and counterpoint here though: the "Didn't Win Title" distribution is very bi-modal. In his last seasons with the Cavs and Heat, his starting teammates' average PER was 15.12 and 13.98, respectively. But in his ensuing first seasons with the Heat and Cavs, his starting teammates' average PER was 18.29 and 18.13, respectively, and they didn't win the title in either of those seasons.

The point I've failed to bring up yet is that none of these numbers exist in a vacuum. Upon reaching the Finals, Jordan faced an arguably lighter set of competition. So I looked at the NBA Finals opponents' starters' average PER to gauge the quality of each year's competition.

SeasonWon TitleOpponent Starters

SeasonWon TitleOpponent Starters
LeBron has faced stiffer competition in the Finals, and the irony is that in his title-winning seasons he actually beat the better teams, on average. His vanquished Finals foes were much better than Jordan's were.

So what can we conclude from this? LeBron did indeed leave when his teammates were weaker, which is a fair criticism when comparing him to Jordan. The thing is, Jordan never had to leave because his teammates were good enough already. And while LeBron is 3-4 in the Finals so far (with 1 pending), he's already been to 8 (and 7 in the last 7 years!!!), whereas Jordan went to 6 and went 6-0 against weaker competition.

Friday, May 26, 2017

Lavar vs Jordan, An Incredibly Lopsided Game of One-on-One

A couple of months ago, in a show of utter brashness and clickbait-worthy quotes, Lavar Ball (father of former UCLA star/lottery pick Lonzo Ball) stated he "would kill Michael Jordan one-on-one". Coming off my post simulating LeBron vs Jordan one-on-one, a friend/colleague brought up that I was looking at the wrong comparison, given Lavar's insane quote. So I threw his stats into the simulator too.

He only played sparingly over one season at Washington State in 1987-88 and averaged 2.2 points a game, so that's all the data I have to go off of. Over 10,000 simulations, Jordan beats him 93.25% of the time by an average score of 21-14.  I'm surprised Lavar's chances are this high, but the simulator doesn't take into account the difference in competition between the NBA and late 80's Pac-10 bench players. Lavar's 40.4% field goal percentage probably wouldn't hold up against the best players on the planet. 

I guess a blind hog finds an acorn roughly 6.75% of the time.

LeBron vs Jordan, One-on-One, in Their Primes

With LeBron reaching his seventh straight NBA Finals, the age-old debate remains: who is better? LeBron or Jordan?

This post isn't looking to answer that question. What it does attempt to answer is this: who would win a game of one-on-one? Last year I created a play-by-play basketball simulator and adapted it to simulate games of pickup, projecting Steph Curry to beat LeBron James in a battle of the best players in the NBA. I've adapted this simulator again, pairing off LeBron and Jordan in their primes in a game of one-on-one.

The first step though is identifying their respective "primes". I compiled the career statistics for each player from Basketball Reference and looked at the best PER seasons for each. I took the top three seasons (per PER) to get metrics such as shooting percentages, rebounding rates, steal rates, etc and averaged them to use as the inputs into the simulator. One thing that is incredibly striking is the eerie similarity for these two GOAT candidates in their top seasons:


Their top 3 PER seasons are exactly the same, at virtually the same ages. 

As before, the pickup rules I used are as follows: 
  1. Standard pickup scoring is applied: 2's and 1's (3-pointers count as 2, 2-pointers count as 1)
  2. No free throws. You don't shoot free throws in pickup
  3. The games are to 21, win by 2
Jordan has a definitive defensive edge, and even though LeBron isn't exactly considered a sharp shooting three-point shooter, he shoots them way more often than Jordan did in his day.

PlayerYearAgePER3P Avg2P Avg3P RatioORebDRebStealBlock
PlayerYearAgePER3P Avg2P Avg3P RatioORebDRebStealBlock

LeBron benefits immensely from the 2's and 1's scoring, and over 10,000 simulations he wins the game of pickup 55.75% of the time by an average score of 21-19.

Friday, May 19, 2017

Efficiency of NCAAF Coaching Salaries, I.E. How Bad Charlie Weis Was

With Notre Dame just now finishing paying out Charlie Weis to NOT coach their football team, much has been made the past few days about how he was ultimately paid $1.6 million per win he achieved between Notre Dame and Kansas. That's $64.5m total for a record of 41-49.

That's really bad! Is it historically bad? I compiled this past year's head coach salaries (gathered from USA Today) and win/loss records to gauge some context.

Weis would have been right at the top of the list of worst pay per win coaches in FBS:

Most Inefficient
VirginiaACCBronco Mendenhall2$1,637,5001
Fresno StateMt. WestTim DeRuyter1$1,518,4802
Michigan StateBig TenMark Dantonio3$1,433,3333
TexasBig 12Charlie Strong5$1,040,0004
RutgersBig TenChris Ash2$1,000,0005
MississippiSECHugh Freeze5$940,0006
MichiganBig TenJim Harbaugh10$900,4007
UCLAPac-12Jim Mora4$862,5008
ArizonaPac-12Rich Rodriguez3$833,3339
OregonPac-12Mark Helfrich4$825,00010

The most inefficient coach, Bronco Mendenhall of Virginia, is almost four times worse than the most efficient coach, Paul Petrino of Idaho:

Most Efficient
IdahoSun BeltPaul Petrino9$46,3451
Appalachian StateSun BeltScott Satterfield10$52,5002
Old DominionC-USABobby Wilder10$55,4593
Louisiana TechC-USASkip Holtz9$55,5564
HawaiiMt. WestNick Rolovich7$57,1445
Eastern MichiganMACChris Creighton7$62,1206
Western MichiganMACP.J. Fleck13$62,7977
TroySun BeltNeal Brown10$66,0008
Southern MississippiC-USAJay Hopson7$71,4299
OhioMACFrank Solich8$71,58910
There aren't any Power 5 conference coaches on this "most efficient" list though. Restricting to these conferences crowns Mike MacIntyre of Colorado as #1:

ColoradoPac-12Mike MacIntyre10$201,1051
WisconsinBig TenPaul Chryst11$245,4552
North CarolinaACCLarry Fedora8$248,2833
West VirginiaBig 12Dana Holgorsen10$298,0004
WashingtonPac-12Chris Petersen12$300,0005
Wake ForestACCDave Clawson7$301,8816
NebraskaBig TenMike Riley9$311,1117
North Carolina StateACCDave Doeren7$314,2868
ClemsonACCDabo Swinney14$315,5369
Virginia TechACCJustin Fuente10$320,00010

A lot of ACC on this list, so you can a lot of mediocrity for your dollar here. For those curious, here's how the top 25 highest paid coaches fair:

1MichiganBig TenJim Harbaugh10$900,400113
2AlabamaSECNick Saban14$495,67186
3Ohio StateBig TenUrban Meyer11$545,72792
4OklahomaBig 12Bob Stoops11$504,54589
5Florida StateACCJimbo Fisher10$525,00090
6TexasBig 12Charlie Strong5$1,040,000116
7Texas A&MSECKevin Sumlin8$625,000103
8AuburnSECGus Malzahn8$590,625100
9MississippiSECHugh Freeze5$940,000114
10IowaBig TenKirk Ferentz8$562,50094
11Penn StateBig TenJames Franklin11$409,09179
12ClemsonACCDabo Swinney14$315,53663
13LSUSECLes Miles8$537,50091
14Michigan StateBig TenMark Dantonio3$1,433,333117
15FloridaSECJim McElwain9$474,25884
16Mississippi StateSECDan Mullen6$700,000108
17ArkansasSECBret Bielema7$585,71498
18TennesseeSECButch Jones9$456,66782
19StanfordPac-12David Shaw10$406,72278
20Texas ChristianBig 12Gary Patterson6$669,121107
21LouisvilleACCBobby Petrino9$431,15981
22Oklahoma StateBig 12Mike Gundy10$377,50072
23GeorgiaSECKirby Smart8$469,20083
24UtahPac-12Kyle Whittingham9$405,55676
25WashingtonPac-12Chris Petersen12$300,00058