A few weeks back, after the game with the Mavericks, I wrote a quick Fanpost noting the statistical similarities between Jonny Flynn's rookie year and Jason Terry's rookie year. However, I wanted to do a bit more analysis before anointing Flynn the next Jason Terry.

Finding a Flynn Comparison

Using Basketball-Reference's season finder, I grabbed every player 6'2" or shorter in their rookie season having played more than 615 minutes played (15mp/g over at least 41 games) since the three point line was put in (1979-1980). I then divided their statistics (MP/G, PER, TS%, eFG%, AST%, STL%, TOV%, USG%, ORtg and DRtg) by Flynn's statistics, giving a percentage result that was, essentially, "closeness to Flynn". I then squared those results and averaged the squares in order to find a comparison to Flynn.

Why Squared?

Pretend that there are only two statistics that matter: TS% and 3P%. Pretend that Flynn has a 50% TS% and 35% 3P%. Player A, has 55% TS% and 31.5% 3P%. Player B, on the other hand, has an 82% TS% and a 12.25% 3P% (impossible, but bear with me). If you do the math, Player A is 10% better at TS%, 10% worse at 3P%. Player B is 65% better and worse at the same statistics. Average them out, and you get 1.00; both are equal in comparison to Flynn. But Player A is obviously the closer comparison. By squaring the results (110% and 90% versus 165% and 35%) and averaging them out, Player A is now a 1.01 in comparison to Flynn's 1.00, while Player B is 1.42. Outliers are made much more significant when squaring.

Weaknesses of This Analysis

First of all, a player with two extreme outliers, even due to squaring, will statistically be closer to Flynn despite the outliers. The raw average (which, BTW, has Mario Chalmers as the closest comparison) is not to be taken at Face value. Chalmers, for instance, was far better at collecting steals, far worse at using possessions and within .2 on just about everything else. Averaging them out mitigates the outliers.

Some statistics are probably more important than others. I'd personally rather have a guy who never steals the ball but shoots 65% from the field. Giving equal weight to everything doesn't make sense in the real world, nor does it make sense statistically. It also makes much more sense to use actual statistical software. When I get a minute, I'll pop in my old SPSS and see if I can't work some more magic.

The biggest problem of all is, strangely, with Flynn. He uses so many possessions, it's astounding. There are SIX short rookies since 1979-1980 who used the ball as much as Flynn - Allen Iverson, Isiah Thomas, Brandon Jennings, Mahmoud Abdul-Rauf, Chucky Atkins and Kelvin Ransey. Jennings is probably the best comparison, and it's not a good one; plus, we don't know what will happen to Jennings in the future.

After the Jump - The Comparisons

Continue reading this post »

In the Minnesota @ Dallas game thread, Ballllin said "i see flynn becoming a terry for us." And while that seems like a pipe dream, their rookie seasons are remarkably similar. Just take a peek at their per/36 statistics as well as some advanced metrics.

The advanced metrics give Terry a slight edge, especially on ORtg and DRtg, but Flynn has more of the pressure on his shoulders, having started the entire season and having the ball in his hands more. He's shooting slightly better despite taking more shots and he's getting to the free-throw line more than Terry did. Having an assist rate lower than Jason Terry doesn't bode well for his future at the point, and it remains to be seen if Flynn can improve like Terry did (especially shooting from the outside) but if Flynn works out, we could have a nice scoring push off the bench.

For the record, the DraftExpress best case for Flynn was Earl Watson, who doesn't match up as well with Flynn's rookie season, but has turned into a mediocre version of the scorer off the bench. That's a pretty bad case scenario for Flynn; Watson is Terry without the great offensive game and there's a reason he's bounced around in his first years as a pro.

Wayne Winston just posted an analysis of the Timberwolves on his blog, Mathletics.

Winston is a big proponent of +/- statistics; he was the expert who argued that Kevin Durant wasn't a positive influence on the Zombie Sonics over the first two years of his career and that Luol Deng is an MVP candidate this year. Agree or disagree with his results or his methods, he has interesting data and interesting conclusions.

Continue reading this post »

Player A, rookie, age 20, per36 minutes - 15pts, 13reb (5 OReb) on 46% shooting, 78% free throw
Player B, rookie, age 21, per36 minutes - 18pts, 11reb (4 OReb) on 55% shooting, 73% free throw

Player A is Kevin Love, player B is Charles Barkley.

Barkley is more well rounded (more steals, assists, blocks) but Love averages fewer turnovers. I say Love could easily become a poor man's Barkley - worse scorer and better rebounder. I wouldn't be surprised to see 18pts, 13reb a game from Love in the very near future. I wouldn't be displeased with that. I'd call Barkley a ceiling for Love and would be very pleased if he became as efficient on the offensive end as Barkley was.