Fall into the Gap: Fantasy Rankings Versus GVT, Forwards
by Bettman's Nightmare on Nov 5, 2010 7:01 AM EDT
Email
PrintReally, this post is an extension of a larger argument between those who are adamant that they were counting what really matters all along and those who have seen hockey analysis as an evolving, academic process. For the moment, the former has dominated the argument when it comes to fantasy hockey, a game that is growing and influencing what fans might define as valuable players.
What I'm doing here is taking two largely-cumulative performance indicators for 2009-10, GVT and Yahoo's fantasy hockey rankings, and seeing which players fall into the gap of disagreement between the two. I'll be taking Yahoo's default scoring system, which attributes positive value to goals, assists, PIMs, power-play points, and shots, and positive or negative value in the case of plus-minus. For a description of GVT, look here. I cut out the forwards who played fewer than half the games of the season to get rid of some of the wonkier numbers, and unfortunately if a player leaves the NHL Yahoo takes them out of their player ranking data from previous years.
As you might expect, a good number of defensive forwards get slighted when moving from GVT to Yahoo, but you also have a number of forwards that created more offense than their numbers indicated. On the flip side, Yahoo gives credit to penalty minutes, a confusing idea, and in the process vaults goon rankings way too high for GVT's liking. Yahoo also will give higher rankings to some of those heavy-offense, questionable-defense-type forwards.
Remember, I'm suggesting that the gap we're seeing here is reflective of an overall gap between what a large number of fans might consider value versus what an advanced metric like GVT (or some in the stats community) would argue.
P.S. Skaters in both categories were ranked across 574 eligible players. Thanks to Behind the Net, data version, for the numbers, as well as Tom Awad.
P.P.S. If you want to know where any other individual player fell, let me know in the comments and I'll look it up.
Largest Gap, GVT Rank Higher Than Yahoo, Top 25 in 2009-10
Rob Schremp 238 GVT Rank | 385 Yahoo Rank | -147 Difference
Jay McClement 209 | 351 | -142
Blair Betts 294 | 432 | -138
Chris Kelly 206 | 323 | -117
Pascal Dupuis 123 | 237 | -114
Lauri Korpikoski 449 | 556 | -107
Sergei Kostitsyn 368 | 473 | -105
Nikolai Kulemin 164 | 266 | -102
Gilbert Brule 160 | 260 | -100
Scott Parse 257 | 350 | -93
Jannik Hansen 411 | 504 | -93
Andrew Murray 478 | 570 | -92
Drew Miller 296 | 387 | -91
Frans Nielsen 139 | 228 | -89
Shawn Matthias 424 | 507 | -83
Michal Handzus 107 | 186 | -79
Pat Dwyer 436 | 515 | -79
Colin Fraser 284 | 359 | -75
Marc Savard 169 | 243 | -74
Ryan Jones 398 | 472 | -74
Daniel Paille 343 | 415 | -72
Kirk Maltby 460 | 531 | -71
Blake Comeau 184 | 250 | -66
Chad LaRose 253 | 318 | -65
Chris Durno 441 | 505 | -64
Average offensive GVT for these players: 1.964 - Average defensive GVT: 1.916
Largest Gap, Yahoo Rank Higher Than GVT, Top 25 in 2009-10
Zenon Konopka 562 GVT Rank | 202 Yahoo Rank | 360 Difference
Colton Orr 549 | 206 | 343
Lee Stempniak 393 | 100 | 293
Bill Guerin 360 | 78 | 282
Ales Kotalik 470 | 193 | 277
Dustin Byfuglien 402 | 128 | 274
Shawn Thornton 550 | 279 | 271
Peter Mueller 498 | 230 | 268
Brandon Prust 484 | 224 | 260
Cody McLeod 522 | 264 | 258
Zack Stortini 502 | 259 | 243
B.J. Crombeen 446 | 204 | 242
Kyle Okposo 326 | 99 | 227
Cam Janssen 555 | 339 | 216
Blake Wheeler 401 | 187 | 214
Patrick O'Sullivan 494 | 283 | 211
Derick Brassard 475 | 265 | 210
Brad Winchester 534 | 327 | 207
Michael Ryder 379 | 174 | 205
Jamal Mayers 491 | 287 | 204
Jared Boll 543 | 343 | 200
Dan Carcillo 325 | 129 | 196
Ruslan Fedotenko 463 | 269 | 194
Scott Hartnell 251 | 64 | 187
Sean Avery 297 | 110 | 187
Average offensive GVT for these players: 0.056 - Average defensive GVT: 0.852
19 comments
|
0 recs |
Do you like this story?
Comments
Are there any fantasy pools that use advanced metrics, like Corsi/Fenwick, GVT, or other?
Using puck possession metrics, which are shot-based, is probably as close as we could come to truly approximating how good a fantasy team would actually have been in reality, from week-to-week.
I haven’t seen any. Yahoo just recently “broke the mold” by incorporating a few new categories, including hits, blocked shots, and faceoff wins. Guy Carbonneau is probably pissed. CBS is the other big game in town, and they use a point system built on the same categories.
I don’t want to rule it out, but I’d say if there is a league using that data, it’s hanging out with the Dodo bird and coelecanths.
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 5, 2010 11:47 AM EDT up reply actions
Yeah, haven’t seen any leagues like that either… I suspect we’ll get there eventually.
SBN Fantasy Hockey Blog: Fantasy Hockey Scouts
by Cam Collingwood on Nov 5, 2010 12:24 PM EDT up reply actions
It’s an interesting list. I’ve thought that Yahoo rankings were getting better year after year, but I see that it’s still not just the goons that are being wildly overrated.
This year Yahoo finally allowed me to replace PIM with Hits in my custom league. It’s not a wonderfully accurate statistic, but it’s better than using ‘how much did you hurt your team’ as a proxy for measuring physical play.
I've been looking at the sky
The thing about Konopka and Orr is that they get a lot of PIM. Since GVT measures a player’s production, PIM doesn’t increase those player’s GVT (unless there’s something I’m missinb about the metric), but in public leagues, where PIM is used, those players become more valuable.
I like to get Matt Carkner because he racks up a lot PIM. What most fantasy hockey leagues won’t show, though, is that Carkner is actually a pretty good defensive player.
As commish of my Yahoo league, every year some manager tries to get PIMs in, and every year I categorically refuse to award points for doing something stupid. We’ve got shots in our league, but I can’t get shots against as a negative to go along with +/- unfortunately. Would be kinda cool.
Anyway, it’s interesting to see that while most of the players undervalued are defensive forwards (as expected), there’s also a solid contingent of offense-first guys that Yahoo thinks are going to fall off a cliff.
I would assume some of that (Savard, for example) would be injury-related.
by SmellOfVictory on Nov 6, 2010 12:49 AM EDT up reply actions
As far as fantasy league vendors that might potentially move to more nerdly measures of a player’s worth, Fantrax ( http://www.fantrax.com ) doesn’t do that yet, but if I were to bet on which site was most likely to move to that first, I’d bet on them. They already incorporate all of the measures that Yahoo does, as well as takeaways, and they’ve got the option to make leagues salary capped, etc.
You’ve convinced me to fight hard to rid my fantasy leagues of PIM’s.
Fantasy Hockey Scouts: a fantasy hockey blog on SB Nation
Tell you what, it’s an uphill battle. I tried to convince the guys in my league to do it, and they basically said they kind of liked PIMs as a value category. I guess there’s always a little admiration for someone who can scored 25-35 goals and still eclipse the 100 PIM mark.
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 6, 2010 3:37 AM EDT up reply actions
If you’re evaluating the value of a player, then counting PIMs as a plus is undeniably stupid. But if you approach a fantasy hockey pool as a game, then including them may lead to a broader and more interesting array of strategic decisions — so long as you realize that at this point you’re just playing with numbers and not with quality, and that a guy who is better for your pool may not be better for your team, and vice-versa.
Therein lies the problem: Yahoo standard scoring makes the decision for normal players by making PIMs one of the default categories. I’m not saying everyone doesn’t have the brains to realized that it’s a negative thing, but it could almost subliminally perpetuate the idea that all those Hartnell minor penalties aren’t such a bad thing.
Hockey’s a bit of a different bird than a lot of the other fantasy sports because there’s a number of categories with debatable real value. What a useless, fluky category shutouts are, for example, and PIMs join a couple of other available categories like hits and game-winning goals that take us nowhere. Goaltender wins have a lot of credibility because there are so few other categories to use for goaltenders; the league I’m in counts saves, just to add a category that might be draftable and to elevate goalies who play well on poorer teams.
In sum, I don’t think it’s safe to assume that fantasy sports, as they are, don’t have an affect on popular notions of the sport and hockey player value. I don’t think all (or even a majority) of fantasy players really separate fantasy value from real value, nor do I think that they think about it that much.
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 9, 2010 3:43 AM EST up reply actions
P.S. Otherwise, you’d probably see Yahoo changing its standard categories. Popular discontent actually did remove game-winning goals from the standard cats, replaced with shots.
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 9, 2010 3:45 AM EST up reply actions
“In sum, I don’t think it’s safe to assume that fantasy sports, as they are, don’t have an affect on popular notions of the sport and hockey player value.”
I agree, but until such time as the sports press doesn’t start doing better analysis (such as using wins as measures of goalie performance without qualification, acquiring an understanding of the functioning of shooting percentages, focussing on scoring chances rather than goals, that sort of thing), it’s an uphill battle anyway and fantasy sports really are just a drop in the bucket. Fans have wrongheaded ideas about the game because those ideas keep being repeated by the “experts”.
TBH, I think evaluating goalies on wins is a worse effect from pool thinking than evaluating players on PIMs. At least with PIMs you can tell someone PIMs are a bad thing and they can see why on some level. Wins are so deeply ingrained in the collective consciousness as the ultimate measure of goaltending that “it’s a result of the team” can never sink in.
I’d probably go a bit further than say that fantasy sports are just a “drop in the bucket.”
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 9, 2010 10:56 AM EST up reply actions
The point is that fans are smart enough to recognize that pool performance != hockey performance, and I think they already do so to some extent. But the sports press uses statistics in a very fantasy-hockey manner which means the average fans just isn’t exposed to better. With the end result that they evaluate players almost exclusively based on goals-assists-points-pims, except grinders, who they evaluate on “grit” and “effort” because that’s what all the ex-grinders-turned-color-men ever talk about.
If the sports press did a better job breaking down things, if NHL plays were broken up to anything near the typical NFL commentary does it (it’s not perfect but damn, they actually break down the strategy and play-calling on a routine basis rather than always talk about effort-effort-effort), then how sports pools work wouldn’t matter as much.
Don’t give fans too much credit: we’re talking about a group that includes young people, people who are new to the game, people who just like to play any fantasy sport they can get their hands on, etc. I can’t tell you the number of people I know that got into hockey because they liked fantasy sports and was hooked into a league started by a friend. This is not to mention we have the “homer” fans, who have difficulty admitting their team is not good regardless of the year.
There’s no doubt that press came first, but there certainly is a give-and-take between press and fantasy nowadays. One is bound to change eventually (press), but I think fantasy hockey would drag its feet on including things like Corsi, or GVT.
Bettman's Nightmare: Hypothesizing that Cooper-alls were the Jorts of the 80s.
http://bettmansnightmare.blogspot.com/
You can also catch my work occasionally at www.behindthenethockey.com
by Bettman's Nightmare on Nov 9, 2010 12:20 PM EST up reply actions
I don’t think fantasy sports will ever go with all-encompassing metrics like GVT on a regular basis for the same reason I’m not sure it makes sense to eliminate PIMs — having different categories allows for more interesting strategic decisions. Fantasy sports are a game first and foremost, after all.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
Related FanPosts on Arctic Ice Hockey
FanPosts
Community blog posts and discussion.
