Any stats for games won/lost when player X plays?

[knorthern knight]

There are stats for games won/lost-regulation/lost-OT-SO by goalies. Are there similar stats for games played by non-goalies? What tweaked my interest was a post by a disgruntled Leafs fan...

Update on the David Clarkson effect:

Leafs without Clarkson 15-6-1 (115 point pace)
Leafs with Clarkson 21-26-7 (74 point pace)

Nazem Kadri without Clarkson: 22gp 8g 17a 25pts +2 (1.14 ppg, 93 point pace)
Nazem Kadri with Clarkson: 49gp 10g 13a 23pts -14 (0.47 ppg, 38 point pace)

I'm not a good screen-scraper. But I run linux at home, and can do a lot if given text or CSV files to work with. So I can stand "raw data" that has to be analyzed+sliced+diced.

Click to read more...

 

[plusandminus]

There are stats for games won/lost-regulation/lost-OT-SO by goalies. Are there similar stats for games played by non-goalies? What tweaked my interest was a post by a disgruntled Leafs fan...



I'm not a good screen-scraper. But I run linux at home, and can do a lot if given text or CSV files to work with. So I can stand "raw data" that has to be analyzed+sliced+diced.

I did this a couple of years ago. I did it for every player on every team, from 1960-61 (or so) to 2010-11 (or so). I posted about it on the forum. Unfortunately the interest was close to zero, and no discussion or debate of value followed. I eventually turned my attention elsewhere, and have been away from here basically since then. The data I had gathered and produced in total took some gigabytes, so I archieved it and deleted in from my computer.

I still do think that this is a very interesting stat, that can be very indicative of a player's worth to hos team. I came to the conclusion that star players like Peter Forsberg, Al MacInnis, Chris Pronger, etc., managed to increase their team's winning % by about 10 percentage units - for example from .500 to .600. Lidstrom had even better stats, but that should not be taken too seriously as Detroit often rested their best players (including Lidstrom) late in the regular season.

Pavol Demitra was a player that often missed many games, and who played for several different teams. He "always" seemed to significantly make his team increase their winning percentage - no matter what season or which team he played for. (Comments on that was that it showed that this is not a useful stat as Demitra was not supposed to be a very good player.)

In 1983-84 Wayne Gretzky scored at least one point in each of Edmonton's 51 first games of the season. He had 153 points in those games, averaging exactly 3.0 points per game. Edmonton was the best team in the league.
With Gretzky (and Kurri) injured, Edmonton then lost 5 out of their next 6 games (including 0-11 against Hartford!), scoring 19 goals and allowing 36.
With Gretzky and Kurri back in the lineup, they immediately went on a 8 game winning streak.
This may not be as indicative as it sound, but I just thought I could mention it.

 

[knorthern knight]

The data I had gathered and produced in total took some gigabytes, so I archieved it and deleted in from my computer.

Forget about the data; do you have the data-gathering routines, or links to how the data can be gathered? I'm insterested in doing that myself.

Pavol Demitra was a player that often missed many games, and who played for several different teams. He "always" seemed to significantly make his team increase their winning percentage - no matter what season or which team he played for. (Comments on that was that it showed that this is not a useful stat as Demitra was not supposed to be a very good player.).

There are some people who disagree with advanced stats, and there are advanced-stats fans who frown on basic stats (like plus and minus ). I'm still interested. This unexpected insight into Demitra's value reminds me of Dave Tippett's famous quote

We had a player that was supposed to be a great, shut-down defenseman. He was supposedly the be-all, end-all of defensemen. But when you did a 10-game analysis of him, you found out he was defending all the time because he can’t move the puck.

Then we had another guy, who supposedly couldn’t defend a lick. Well, he was defending only 20 percent of the time because he’s making good plays out of our end. He may not be the strongest defender, but he’s only doing it 20 percent of the time. So the equation works out better the other way. I ended up trading the other defenseman.

 

[plusandminus]

Forget about the data; do you have the data-gathering routines, or links to how the data can be gathered? I'm insterested in doing that myself.

One needs game stats from all the 1230 games during the season. I probably downloaded them from nhl.com, then wrote a program who converted them to something a computer could understand, and then I stored the data into a SQL Server database.
After that, I wrote code that handled the algoritm of finding out how the team performed with or without the player.
It's much easier with recent seasons. For older seasons, very much time and effort was directed into finding and correcting errors in the data (miss-spelled names, inconsistant formatting, etc.).

There are some people who disagree with advanced stats, and there are advanced-stats fans who frown on basic stats (like plus and minus ). I'm still interested. This unexpected insight into Demitra's value reminds me of Dave Tippett's famous quote

Yes. For example, there was some criticism towards the Sedin twins when they were young, for their supposed lack of defensive ability. Many counter argued that the best way to defend is to let the puck be in the opposing end (like the Sedins did), as few goals are being given up from that area.
I agree that +/- in itself is not a very reliable stat, and it tends to change so much from season to season. (My user name rather is about seing both sides of things in general.) Ice time and points scored are much more reliable, even though scoring of course is a one-dimensial stat.