Help apreciated...

[CollegeKingRex]

Sup fellas, hope all is well in here.

As always, I'm keeping my eyes open from the sidelines regarding the hockey season. You'll see me around playoff time, as usual.

Trying to put in some input to a few smart friends here (they are stone broke, lmao) trying to build a hockey database. They (correctly, I might add), are trying to build a database that will help spit out strong hockey lines, trends etc. to sell to a guy in Vegas looking to buy it before the start of next season. He won't likely be betting the games until 2017 so these guys will have like 9 months to build their machine, and they enlisted my help (my cost is receiving the plays once the machine starts spitting them out, which I will gladly share with the forum once I get down - i won't be betting enough on them to move the market most likely)

Anyway, these guys want to look at the key stats when trying to figure out how to build their box. Even if it's only the top 10 or 15, if y'all can post in here or send me a PM of your top 10-20 stats - both offensive and defensive, it would be appreciated.

I have my own opinions (No. 1 is shots on goal, and No. 1A is opponents shots on goal, for example), but I trust the small but stout CTG hockey community, which I respect immensely, more than I do myself. All help appreciated in advance, and within the next couple of days if you've got the time to respond to me.

:shake:

 

[Lareux]

Corsi # - this stat takes all shots taken, not just those on net (blocked, missed net, etc)...it is much more indicative of a team's success/failure than simply shots on goal

There are a few other advanced stats like this that should be helpful. Let me give a shout to a friend of a friend (who parlayed his analysis of the Sabres farm system into a job with the team), and see some of the stats he recommends for this project.

:shake:

 

[CollegeKingRex]

ah yes, CORSI... similar to KUBIAK for the guys who write the Hockey/Football Outsiders books. Must reads every year before betting any of these games (although in football this year, I rated Ariz and Tampa much higher than them, and don't always blindly agree with them)

 

[CollegeKingRex]

also larry, i need to see what they can write the code off of, but i'm not sure CORSI is even one of the options.

Let me post what they can scoop up off the xml feed in here. I'll have him send it to me.

 

[Lareux]

also larry, i need to see what they can write the code off of, but i'm not sure CORSI is even one of the options.

Let me post what they can scoop up off the xml feed in here. I'll have him send it to me.

Sounds good. I'd imagine they may be able to scoop that Corsi # as so many people are now talking about it. It's quite possible that box scores are now showing ALL shot attempts, not just those on net.

 

[CollegeKingRex]

i heard from a canadian friend of mine here that they don't value CORSI as much as y'all are led to believe, that Fenwick is more reliable.

 

[CollegeKingRex]

Here are the categories they can draw from.

Please and thanks to anyone who chimes in on this in here and/or to my PM.

:shake:


[TABLE="width: 292"]
<colgroup><col></colgroup><tbody>[TR]
[TD] attendance_total[/TD]
[/TR]
[TR]
[TD] conference_ranking[/TD]
[/TR]
[TR]
[TD] division_ranking[/TD]
[/TR]
[TR]
[TD] faceoff_total_losses[/TD]
[/TR]
[TR]
[TD] faceoff_total_wins[/TD]
[/TR]
[TR]
[TD] games_lost[/TD]
[/TR]
[TR]
[TD] games_lost_overtime[/TD]
[/TR]
[TR]
[TD] games_lost_shootout[/TD]
[/TR]
[TR]
[TD] games_played[/TD]
[/TR]
[TR]
[TD] games_tied[/TD]
[/TR]
[TR]
[TD] games_won[/TD]
[/TR]
[TR]
[TD] games_won_overtime[/TD]
[/TR]
[TR]
[TD] games_won_shootout[/TD]
[/TR]
[TR]
[TD] goalie_assists[/TD]
[/TR]
[TR]
[TD] goalie_assists_against[/TD]
[/TR]
[TR]
[TD] goalie_goals_against[/TD]
[/TR]
[TR]
[TD] goalie_goals_against_average[/TD]
[/TR]
[TR]
[TD] goalie_goals_against_average_against[/TD]
[/TR]
[TR]
[TD] goalie_goals_allowed_against[/TD]
[/TR]
[TR]
[TD] goalie_goals_scored[/TD]
[/TR]
[TR]
[TD] goalie_goals_scored_against[/TD]
[/TR]
[TR]
[TD] goalie_hits[/TD]
[/TR]
[TR]
[TD] goalie_losses_against[/TD]
[/TR]
[TR]
[TD] goalie_penalty_minutes[/TD]
[/TR]
[TR]
[TD] goalie_penalty_minutes_against[/TD]
[/TR]
[TR]
[TD] goalie_points_power_play[/TD]
[/TR]
[TR]
[TD] goalie_points_short_handed[/TD]
[/TR]
[TR]
[TD] goalie_saves[/TD]
[/TR]
[TR]
[TD] goalie_saves_against[/TD]
[/TR]
[TR]
[TD] goalie_shots_against[/TD]
[/TR]
[TR]
[TD] goalie_time_on_ice_power_play_secs[/TD]
[/TR]
[TR]
[TD] goalie_time_on_ice_secs[/TD]
[/TR]
[TR]
[TD] goalie_time_on_ice_secs_against[/TD]
[/TR]
[TR]
[TD] goalie_time_on_ice_short_handed_secs[/TD]
[/TR]
[TR]
[TD] goalie_wins_against[/TD]
[/TR]
[TR]
[TD] goals[/TD]
[/TR]
[TR]
[TD] goals_against[/TD]
[/TR]
[TR]
[TD] goals_empty_net[/TD]
[/TR]
[TR]
[TD] goals_empty_net_against[/TD]
[/TR]
[TR]
[TD] goals_overtime[/TD]
[/TR]
[TR]
[TD] goals_overtime_1[/TD]
[/TR]
[TR]
[TD] goals_overtime_1_against[/TD]
[/TR]
[TR]
[TD] goals_overtime_2[/TD]
[/TR]
[TR]
[TD] goals_overtime_2_against[/TD]
[/TR]
[TR]
[TD] goals_overtime_3[/TD]
[/TR]
[TR]
[TD] goals_overtime_3_against[/TD]
[/TR]
[TR]
[TD] goals_overtime_against[/TD]
[/TR]
[TR]
[TD] goals_period_1[/TD]
[/TR]
[TR]
[TD] goals_period_1_against[/TD]
[/TR]
[TR]
[TD] goals_period_2[/TD]
[/TR]
[TR]
[TD] goals_period_2_against[/TD]
[/TR]
[TR]
[TD] goals_period_3[/TD]
[/TR]
[TR]
[TD] goals_period_3_against[/TD]
[/TR]
[TR]
[TD] goals_power_play[/TD]
[/TR]
[TR]
[TD] goals_power_play_against[/TD]
[/TR]
[TR]
[TD] goals_shootout[/TD]
[/TR]
[TR]
[TD] goals_shootout_against[/TD]
[/TR]
[TR]
[TD] penalties[/TD]
[/TR]
[TR]
[TD] penalty_minutes[/TD]
[/TR]
[TR]
[TD] player_assists[/TD]
[/TR]
[TR]
[TD] player_assists_against[/TD]
[/TR]
[TR]
[TD] player_blocked_shots[/TD]
[/TR]
[TR]
[TD] player_game_winning_goals[/TD]
[/TR]
[TR]
[TD] player_game_winning_goals_against[/TD]
[/TR]
[TR]
[TD] player_hits[/TD]
[/TR]
[TR]
[TD] player_penalty_minutes_against[/TD]
[/TR]
[TR]
[TD] player_plus_minus[/TD]
[/TR]
[TR]
[TD] player_plus_minus_against[/TD]
[/TR]
[TR]
[TD] player_points[/TD]
[/TR]
[TR]
[TD] player_points_against[/TD]
[/TR]
[TR]
[TD] player_points_power_play[/TD]
[/TR]
[TR]
[TD] player_points_short_handed[/TD]
[/TR]
[TR]
[TD] player_time_on_ice_power_play_secs[/TD]
[/TR]
[TR]
[TD] player_time_on_ice_short_handed_secs[/TD]
[/TR]
[TR]
[TD] power_plays[/TD]
[/TR]
[TR]
[TD] power_plays_against[/TD]
[/TR]
[TR]
[TD] shootouts_attempted[/TD]
[/TR]
[TR]
[TD] shootouts_attempted_against[/TD]
[/TR]
[TR]
[TD] shootouts_scored[/TD]
[/TR]
[TR]
[TD] shootouts_scored_against[/TD]
[/TR]
[TR]
[TD] shots[/TD]
[/TR]
[TR]
[TD] shots_against[/TD]
[/TR]
[TR]
[TD] shots_overtime[/TD]
[/TR]
[TR]
[TD] shots_overtime_1[/TD]
[/TR]
[TR]
[TD] shots_overtime_1_against[/TD]
[/TR]
[TR]
[TD] shots_overtime_2[/TD]
[/TR]
[TR]
[TD] shots_overtime_2_against[/TD]
[/TR]
[TR]
[TD] shots_overtime_3[/TD]
[/TR]
[TR]
[TD] shots_overtime_3_against[/TD]
[/TR]
[TR]
[TD] shots_overtime_against[/TD]
[/TR]
[TR]
[TD] shots_period_1[/TD]
[/TR]
[TR]
[TD] shots_period_1_against[/TD]
[/TR]
[TR]
[TD] shots_period_2[/TD]
[/TR]
[TR]
[TD] shots_period_2_against[/TD]
[/TR]
[TR]
[TD] shots_period_3[/TD]
[/TR]
[TR]
[TD] shots_period_3_against[/TD]
[/TR]
[TR]
[TD] shutouts[/TD]
[/TR]
[TR]
[TD] shutouts_against[/TD]
[/TR]
[TR]
[TD] total_play_time[/TD]
[/TR]
</tbody>[/TABLE]

 

[Lareux]

i heard from a canadian friend of mine here that they don't value CORSI as much as y'all are led to believe, that Fenwick is more reliable.

Could be, Fenwick would be another one on the list. I was just saying Corsi is more reliable/valuable than shots on goal. You may very well be right about Fenwick being more reliable than Corsi. For this exercise, it probably makes sense to use both Corsi and Fenwick anyway.

 

[CollegeKingRex]

Well, looks I'm shit outta luck when it comes to the analytic stuff. The guy says that everything needs come from the list of stuff I just posted above in post 7.

Any and all help still appreciated here, y'all...

 

[guaranteeed]

is there no way they can pull or scrub data from war-on-ice or something like it?

I would say Fenwick>Corsi as well, just because it excludes blocked shots which is a repeatable skill and not luck (both ways, some teams excel at it, some players excel at getting shots past it)

but in general things that are widely available like Fenwick or Corsi 5on5 %, F/C 5on5 score close % are probably the two best as they will take out score effects from muddying the data

PDO is good to show who has been luck and is due to regress, either good or bad


but in terms of whats there

shots for
shots against
shorts for minus shots against

are the best 3, they are going to show similar stats to Corsi/Fenwick, just over the season the sample size will be a lot smaller, and score effects, powerplay time etc. will make it not as valuable

you could create your own PDO

need to pull goals and divide by shots, plus goals against divided by shots against. League Average will end up at 1.00 and teams will regress to the mean. This is a really good one I think to target teams getting unlucky or lucky

last year all but 6 teams finished between 0.99 and 1.01, so looking at this year a team like the Rangers is at 1.04, they are due to regress. The one caveat with that is teams with elite goalies are usually due an extra 0.01 just due to that extra save percentage, but that is limited to maybe the top 2/3 guys in the year

 

[Lareux]

I would say Fenwick>Corsi as well, just because it excludes blocked shots which is a repeatable skill and not luck (both ways, some teams excel at it, some players excel at getting shots past it)

Since blocked shots are a repeatable skill, any reason it would be better to exclude it? I'm just looking at it similar to a save by a goalie, seems it would be more accurate to include the blocked shots. I know some guys see Fenwick as more valuable, which does exclude blocked shots, I'm just wondering the logic behind it. I'm missing it for sure.

 

[Lareux]

Found a quote directly from Matt Fenwick on the topic...

"My argument is basically:
1. The whole (or perhaps best) use of Corsi is to have objective figures that can be used as a proxy for scoring chances (what else are you using it for?).
2. A shot that is blocked is either a) not a scoring chance at all, or b) on average from a worse scoring area than shots/posts/missed shots.
Yes it affects the "sample size" but that only means anything if what you are sampling is relevant to what you are trying to represent. You could include Penalties Drawn (or hell, faceoff wins/losses!), but I'm not sure that the connection to a scoring chance is as obvious as for shots/posts etc.
That's my opinion anyway, take it or leave it."
- Matt Fenwick, Nov. 22nd, 2007.



​

 

[Lareux]

Makes more sense now, he's really just trying to isolate scoring chances. The only argument against it would be since blocked shots are a skill, the blocked shots could, in fact, have been a scoring chance had the defenceman's skill not prevented the shot from getting through. Both stats (Corsi, Fenwick) seem to have their place though, and it seems clear from the data that Fenwicks is better at determining future outcomes...so perfect for this exercise Rexy.