Monday 2 January 2012

Putting Some Numbers on Hobey

I've written about Hobey Baker before, more specifically whether he belongs in the Hall of Fame. A quick summary goes something like this: he was a very good player, had superficially impressive numbers, but due to the low level of competition he played against, that's not enough to merit induction into the Hall of Fame. His tragic death is likely what got him in. Also, the lack of quality competition means that people had only reports to know how good he was; he never faced the best players of his time which might have revealed the flaws in his game that were not apparent when opposing lesser teams.

And Baker did have flaws. Although he was a gifted shooter, passer, and (especially) skater, possessing blinding speed, he was also a floater. Descriptions of his play make it clear that he did very little backchecking, instead waiting for his defencemen to recover the puck from the opponents and get the offence started so Baker could make another of his dandy rushes. Now, being purely an offensive player does not make you a bad player, as I mentioned when discussing Harry Smith. But the existing writing on Baker rarely implies anything other than his perfection as a player. He has been heroicized to the extent that objective discussion of his abilities as a hockey player is difficult.

One of the reasons for the development of the historical Point Allocation system is to allow us to compare players from different times and leagues to estimate how much they contributed to their teams' success, and including an adjustment for a quality of the league. The league quality adjustment allows us, in theory, to produce Point Allocation numbers for any player in any league. This means we can determine numbers for Hobey Baker, not only for his two seasons in the American Amateur Hockey League (AAHL), but also his three years at Princeton. So the new year bring the first look at a player's career Point Allocation results:

SeasonTeam (League)PosGPMINOPDPPPTPATPAK
1912Princeton Tigers (NCAA)480160010.0-3.30.06.74.19
1913Princeton Tigers (NCAA)48016005.6-2.70.02.91.81
1914Princeton Tigers (NCAA)48016007.1-2.20.04.93.06
1915New York St. Nicks (AAHL)480160010.8-2.90.07.94.94
1916New York St. Nicks (AAHL)47314602.4-0.70.01.71.16

Point Allocation results are normalized to an 80-game schedule, and forwards are normalized to 20 minutes per game played. The table above show Baker's Offensive Points (OP), Defensive Points (DP), Penalty Points (PP), and Total Points Allocated (TPA). The final column, TPAK, is TPA per 1,000 minutes, which is used to compare the performances of players at different positions.

Of course, at this time you have no point of reference for the above numbers. Is a TPAK of 4.94 impressive? If so, how impressive is it? To get an idea, we have to look at results for more recent players, for whom we have better ideas of how good they were. So, here are the top 30 players for the 1967/68 NHL season, by TPAK:

PlayerTeamPosGPMINOPDPPPTPATPAK
Beliveau, JeanMTL56412806.61.4-0.17.96.17
Gamble, BruceTORG5733640.021.00.021.05.93
Bower, JohnnyTORG5834220.020.70.020.75.85
Tremblay, J.C.MTL37919752.29.30.011.55.82
Esposito, PhilBOS58016007.31.3-0.18.55.31
Ratelle, JeanNYR58016007.11.4-0.18.45.25
Cournoyer, YvanMTL76913805.71.4-0.17.05.07
Laperriere, JacquesMTL37819501.68.5-0.39.85.03
Neilson, JimNYR37218003.06.3-0.39.05.00
Goyette, PhilNYR57915806.61.3-0.17.84.94
Williams, TommyBOS57414806.41.0-0.17.34.93
Bucyk, JohnnyBOS67815606.60.90.07.54.81
Gilbert, RodNYR77915806.31.20.07.54.75
Howe, GordieDET78016006.71.0-0.27.54.69
Brown, ArnieNYR38020002.27.2-0.39.14.55
Tremblay, GillesMTL67715404.03.00.07.04.55
Howell, HarryNYR38020002.17.2-0.39.04.50
Rousseau, BobbyMTL78016005.51.6-0.17.04.38
Favell, DougPHIG5633500.015.70.015.74.37
Mikita, StanCHI57815605.31.50.06.84.36
Harris, TedMTL37218002.26.0-0.47.84.33
Hodge, KenBOS78016006.10.8-0.16.84.25
Provost, ClaudeMTL77915803.92.9-0.16.74.24
Hull, BobbyCHI67715405.01.7-0.26.54.22
Stanfield, FredBOS67915805.80.80.06.64.18
McKenzie, JohnBOS78016006.20.9-0.56.64.13
Delvecchio, AlexDET58016005.51.2-0.16.64.13
Ullman, NormDET56312604.70.50.05.24.13
Prentice, DeanDET67515005.31.0-0.26.14.07
Green, TedBOS37819503.94.3-0.47.84.00

So a TPAK of 4.94 is close to the level of being the very best in the NHL, which is impressive indeed. If Baker had been able to produce at that level for his career, he would have a strong argument for the Hall of Fame, even with his short career. But he didn't, of course. His second-best season was 4.19, which is still plenty good. However his third-best was only 3.06 (1967/68 equivalents: Earl Ingarfield or Ed Westfall), the level of a very solid NHLer but not a star. Then we get 1.81 (Claude Laforge or Pat Hannigan), and 1.16 (Billy Harris or Wally Boyer), which are not numbers recorded by Hall of Famers.

Especially when examining players of Baker's era, it's important to look at their career results rather than just a single season. Since they played such a low number of games per season, it's much easier for normal variance in performance to creep its way into seasonal results (although it is easy to overstate the importance of the number of games played; Baker may have only played eight games per year, but he was playing 60 minutes per game, and the minutes are really what matters for variance purposes). Baker's average season was just over 3.00 TPAK, meaning that if he did play in the NHL he would have been a very good player, contributing a lot to his team's success. But he would not be one of the very best players, not one who you would expect to see in the Hall of Fame when his playing days were done.

No disrespect intended to the man or his legacy, of course. But no one should be above rational analysis. Even heroes are subject to reality checks.

P.S. Here are renditions of the sweaters Baker wore during his career. They don't add anything to our knowledge of Baker, but they're fun.

No comments:

Post a Comment

Hostgator promo codes