Most worthless stat in sports?

[Suburban Rhythm]

Thought about making this a poll, but didn't want to limit the answers to only those choices I came up with.

The 'NHL removing divisions' thread as well as a conversation at work yesterday had me thinking about this.

I've got a few answers bouncing around my head, but for me it's the GWG in hockey. Scoring the 2nd goal in what becomes a 5-0 lead for your team, before surrendering 1 goal with minutes to play, doesn't make your goal all that important, and certainly not enough to have a stat made up to recognize said goals.

[cody8200]

A stathead would definitely say Wins and Losses for baseball. Both of which the pitcher has little to do with.

[JediKooter]

A stathead would definitely say Wins and Losses for baseball. Both of which the pitcher has little to do with.

So Barry Zito didn't have much to do with his win/loss record the last few years?

[cody8200]

So Barry Zito didn't have much to do with his win/loss record the last few years?

I should have just said wins. Losses are much more controllable.

More details:

http://insider.espn.go.com/mlb/insider/columns/story?columnist=law_keith&id=5599365&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fmlb%2finsider%2fcolumns%2fstory%3fcolumnist%3dlaw_keith%26id%3d5599365

[LastWhiteSoxFanStanding]

A quality start is 6 IP 3 ER which equals a 4.50 era.

[Rizon]

A quality start is 6 IP 3 ER which equals a 4.50 era.

Below average is the new quality! :lol:

[larrymcg421]

Saves in baseball. Not only is it useless, but it has been damaging to the strategy of the game.

[gstelmack]

Agreed on saves in baseball. What about WAR? So useless no one can agree on a definition!

[Rizon]

Agreed on saves in baseball. What about WAR? So useless no one can agree on a definition!

What is it good for? Absolutely nothing.

[Marc Vaughan]

Pretty much any soccer stat which isn't taken in context with other information - for instance a team might have 60% posession because the opposition is playing counter attacking and allowing them to push forward or because they are truly dominating a game ....

Similarly a team might have had 20 shots on goal all from 30 yards or might have had them all from close range, context is everything.

[Pumpy Tudors]

Agreed on saves in baseball. What about WAR? So useless no one can agree on a definition!
don't talk shit about war
we are going to start using that to decide the mvp

[Logan]

BCS Average.

[JediKooter]

I should have just said wins. Losses are much more controllable.

More details:

Keith Law: Starting pitchers only responsible for piece of the 'Pie' - ESPN (http://insider.espn.go.com/mlb/insider/columns/story?columnist=law_keith&id=5599365&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fmlb%2finsider%2fcolumns%2fstory%3fcolumnist%3dlaw_keith%26id%3d5599365)

Ah gotcha.

[Toddzilla]

whatever that horseshit QB rating stat ESPN conjured up this summer

[A-Husker-4-Life]

whatever that horseshit QB rating stat ESPN conjured up this summer

this

[korme]

A stathead would definitely say Wins and Losses for baseball. Both of which the pitcher has little to do with.

Thread over

[spleen1015]

Comparing how good a QB is and how much his team wins....

[korme]

Just wins, though. As mentioned. A pitcher doing a great job at sucking deserves all the L's he racks up

[Simbo Klice]

Red Zone Scoring %. Since a FG counts for the same as a TD, it's usually misleading; they should redo it to be points per trip.

[Dr. Sak]

+/- in Hockey

[Coffee Warlord]

Baseball Wins, hands down.

Though I have to agree, ESPN really outdid themselves with that horrid QB rating they came up with.

And on a similar subject...most QB based stats in general are garbage. I'd really like to see some sort of advanced metric for QB completions, factoring in drops & Blatantly Not The QB's Fault picks.

[Grover]

+/- in Hockey

This, by far.

[Izulde]

What makes +/- worthless in hockey? The lower number of points scored? It's a genuine question, as I find it of some value in basketball.

[Autumn]

Baseball Wins, hands down.

Though I have to agree, ESPN really outdid themselves with that horrid QB rating they came up with.

And on a similar subject...most QB based stats in general are garbage. I'd really like to see some sort of advanced metric for QB completions, factoring in drops & Blatantly Not The QB's Fault picks.

Have you looked at Football Outsider QB stats? I feel like they do an admirable job of looking at the real picture as much as they can.

[jbergey22]

And on a similar subject...most QB based stats in general are garbage. I'd really like to see some sort of advanced metric for QB completions, factoring in drops & Blatantly Not The QB's Fault picks.


FOOTBALL OUTSIDERS: Innovative Statistics, Intelligent Analysis | 2011 QUARTERBACKS (http://www.footballoutsiders.com/stats/qb)

[jbergey22]

Many great opinions so far.

Im going to say turnovers in basketball. As with many of the suggested ones, this on its own says nothing as a stat. In most years the 5-10 best players in the NBA are on this leaderboard because they have the ball in their hands so much and are making things happen.

As a team stat it means something, individually I dont feel it means much of anything.

[jbergey22]

What makes +/- worthless in hockey? The lower number of points scored? It's a genuine question, as I find it of some value in basketball.

Im not a big hockey guy but I assume they mean with the different lines out there that do different things. You may play your best defensive line against the other teams best offensive line which will make your +/- look poor on that defensive line.

[korme]

+/- in basketball is really kind of stupid, as well.

[albionmoonlight]

This might not be quite at the level of some of the others, but I'd say rushing touchdowns in football. That, to me, has much more to do with opportunity than ability.

Also, tackles in football are a pretty useless stat. Especially because hometown scorekeepers freely assign tackles to the team's defensive star as long as he was within 3 yards of the ballcarrier when he went down.

[gstelmack]

Red Zone Scoring %. Since a FG counts for the same as a TD, it's usually misleading; they should redo it to be points per trip.

Yards gained / yards allowed, used as metrics for ranking offenses and defenses.

Points scored / points allowed is much closer, although we need to factor defensive TDs in here somewhere properly, they aren't "scored" by the offense or "allowed" by the defense.

[Grover]

What makes +/- worthless in hockey? The lower number of points scored? It's a genuine question, as I find it of some value in basketball.

Hockey Prospectus | Numbers On Ice: Fixing Plus/Minus (http://www.puckprospectus.com/article.php?articleid=64)

Pretty good article summing it up. Good players on bad teams are going to have more ice time and likely a higher minus number because of it. Plus/Minus is not indicative of skill or talent whatsoever.

[korme]

Yards gained / yards allowed, used as metrics for ranking offenses and defenses.

Also agreed. Football is still pretty much in the stoneage when it comes to modern statistical analysis.

[Suburban Rhythm]

+/- in Hockey

This, by far.

Hockey Prospectus | Numbers On Ice: Fixing Plus/Minus (http://www.puckprospectus.com/article.php?articleid=64)

Pretty good article summing it up. Good players on bad teams are going to have more ice time and likely a higher minus number because of it. Plus/Minus is not indicative of skill or talent whatsoever.

I think the way most people use/perceive it, it's a terrible stat. But used "properly", it's not bad. Not great, but not bad. Taken in context with other information like QUALCOMP, it can be helpful.


As I said in my original post, GWG is crap. Its tells you nothing, other than a guy scored A goal at some point in A game...which possibly 50+ minutes later becomes the goal the is the one for his team that just so happens to be 1 more than the other team scored.

[Suburban Rhythm]

This might not be quite at the level of some of the others, but I'd say rushing touchdowns in football. That, to me, has much more to do with opportunity than ability.

Also, tackles in football are a pretty useless stat. Especially because hometown scorekeepers freely assign tackles to the team's defensive star as long as he was within 3 yards of the ballcarrier when he went down.

The Ray Lewis rule. Even if he's the 5th guy on the pile, tackle.

Hits in hockey are notoriously skewed in certain rinks.

[Grover]

Suburban, how do you feel about CORSI?

[DanGarion]

Any statistic that includes a number instead of a subjective explanation of players perceived ability and skill.

[bulletsponge]

Baseball Wins, hands down.

Though I have to agree, ESPN really outdid themselves with that horrid QB rating they came up with.

And on a similar subject...most QB based stats in general are garbage. I'd really like to see some sort of advanced metric for QB completions, factoring in drops & Blatantly Not The QB's Fault picks.

+1000000. passing yards and passing TD's are touted every week but indicate nothing on how well he played. worthless stats

[Crapshoot]

RBI's are clearly horrific, and wins for pitchers are pretty mediocre (still, better than RBI's).

[RomaGoth]

I think the way most people use/perceive it, it's a terrible stat. But used "properly", it's not bad. Not great, but not bad. Taken in context with other information like QUALCOMP, it can be helpful.


As I said in my original post, GWG is crap. Its tells you nothing, other than a guy scored A goal at some point in A game...which possibly 50+ minutes later becomes the goal the is the one for his team that just so happens to be 1 more than the other team scored.

I was under the impression that a goal only counts as a GWG if it is a 2 goal or less victory. In your example, a 5-1 win would not qualify that 2nd goal as a GWG. Unless something has changed with how it is measured or I am mistaken (which would not be unusual in the least).

[Grover]

I was under the impression that a goal only counts as a GWG if it is a 2 goal or less victory. In your example, a 5-1 win would not qualify that 2nd goal as a GWG. Unless something has changed with how it is measured or I am mistaken (which would not be unusual in the least).

A game-winning goal (GWG) is the (n + 1)th goal scored by a team that defeats a team which scores n goals; e.g. if team A beats team B 2-1, the scorer of team A's second goal is credited with the GWG, if team A beats team B 4-2, the scorer of team A's third goal is credited with the game-winner.

[larrymcg421]

I was under the impression that a goal only counts as a GWG if it is a 2 goal or less victory. In your example, a 5-1 win would not qualify that 2nd goal as a GWG. Unless something has changed with how it is measured or I am mistaken (which would not be unusual in the least).

So then you change the example to the 4th goal in a 5-3 win. It's still a stupid stat. Baseball used to have a similarly stupid stat GWRBI and wisely scrapped it for the same arguments made above. It says quite a bit when baseball, king of the ridiculous stats, gets rid of a stat because it is a stupid.

[johnnyshaka]

Just wins, though. As mentioned. A pitcher doing a great job at sucking deserves all the L's he racks up

So a pitcher losing 1-0 deserves the L? A pitcher loses 10-0 but all the runs are unearned and he still deserves the L?

[SteveMax58]

Balks :D

[albionmoonlight]

Oh, and any stat where the cut-off number is clearly arbitrary and created for purposes of the stat.

I am seeing this a lot with Drew Brees right now. He's having an epic run, but some of the "records" that they are talking about are not "records" in any real sense. Things like consecutive games with 20+ completions. Why 20? Because he probably had less than 25 in a few of those games. And some other guy would probably have the record if you made it 18.

Another example is something like "Only player to go 5 or more games in a row with at least 17 points, 10 rebounds, and 7 assists."

Even though all numbers are somewhat arbitrary, I get using certain agreed upon milestone numbers (100 yards rushing, hat trick of goals, 10 rebounds, etc.) to base a "record" on. It is when you don't even bother to tether yourself to that that I have a problem.

[RomaGoth]

A game-winning goal (GWG) is the (n + 1)th goal scored by a team that defeats a team which scores n goals; e.g. if team A beats team B 2-1, the scorer of team A's second goal is credited with the GWG, if team A beats team B 4-2, the scorer of team A's third goal is credited with the game-winner.

I understand how the stat works, I just wasn't aware that it is still considered a GWG if a team wins by 3 or more goals.

So then you change the example to the 4th goal in a 5-3 win. It's still a stupid stat. Baseball used to have a similarly stupid stat GWRBI and wisely scrapped it for the same arguments made above. It says quite a bit when baseball, king of the ridiculous stats, gets rid of a stat because it is a stupid.

Oh I totally agree the stat is worthless.

[Groundhog]

POS MIN FG-FGA FT-FTA 3P-3PA ORB-TRB AS ST BL TO PF TP
Earl,Acie C 44 13-23 14-19 0- 1 4-12 3 0 2 6 3 40

[Maple Leafs]

A game-winning goal (GWG) is the (n + 1)th goal scored by a team that defeats a team which scores n goals; e.g. if team A beats team B 2-1, the scorer of team A's second goal is credited with the GWG, if team A beats team B 4-2, the scorer of team A's third goal is credited with the game-winner.
Only semi-related, but a good example of ridiculous stats.

Last night, Carolina brings in goalie Mike Murphy late in the third period while trailing Calgary 6-3. It's 6-4 when he ends up getting pulled for an extra attacker with a minute left. The Flames score an empty netter to make it 7-4. Carolina then almost pulls off the miracle comeback, scoring twice in the final minute. Final score is Calgary 7, Carolina 6.

The end result: Because of the comeback, the Flames' seventh goal (into an empty net) becomes the "winner". And because Murphy was the goalie of record, he takes the loss even though he was on the bench when it was scored.

The kicker is that it was his NHL debut, making him the first (and let's just assume only) goalie in NHL history to lose a game before ever giving up a goal.

[RomaGoth]

+/- in Hockey

Slightly worse than GWG. The +/- stat in hockey measures a player individually but is really an indication of the entire line of players on the ice, not just that player's individual performance.

I hate +/-.

:rant:

[Suburban Rhythm]

Suburban, how do you feel about CORSI?

I think it helps confirm notions most people feel, but normally can't quantify. Zetterberg is typically at or near the top of this list. Most people can watch a game and see he, and his team, posses the puck and spend a lot of time in the opponent's zone, but now have a concrete number.

[Suburban Rhythm]

Slightly worse than GWG. The +/- stat in hockey measures a player individually but is really an indication of the entire line of players on the ice, not just that player's individual performance.

I hate +/-.

:rant:

While I will agree it is flawed, itself better than GWG.

The flaw with +/- is using it across teams, rather than comparing players on the same team.

[britrock88]

Only semi-related, but a good example of ridiculous stats.

Last night, Carolina brings in goalie Mike Murphy late in the third period while trailing Calgary 6-3. It's 6-4 when he ends up getting pulled for an extra attacker with a minute left. The Flames score an empty netter to make it 7-4. Carolina then almost pulls off the miracle comeback, scoring twice in the final minute. Final score is Calgary 7, Carolina 6.

The end result: Because of the comeback, the Flames' seventh goal (into an empty net) becomes the "winner". And because Murphy was the goalie of record, he takes the loss even though he was on the bench when it was scored.

The kicker is that it was his NHL debut, making him the first (and let's just assume only) goalie in NHL history to lose a game before ever giving up a goal.

This is wonderful. The only other scenario I can think of where that could happen is a first-appearance goalie who carries a shutout into the shootout and loses there.

[Logan]

We talked about this in the MLB thread, but virtually every "most career/4th highest/top 10 XXs in postseason history" stat is worthless by way of the changes in the number of potential games played over the years.

[BYU 14]

Quality starts in Baseball for sure, time of posession in Football, +/- in hockey and laps led in racing.

[Logan]

Quality starts in Baseball for sure, time of posession in Football, +/- in hockey and laps led in racing.

Thread title says "sports".

[korme]

So a pitcher losing 1-0 deserves the L? A pitcher loses 10-0 but all the runs are unearned and he still deserves the L?

those are not the case of a pitcher sucking, the great adjective i used to describe that situation

[korme]

POS MIN FG-FGA FT-FTA 3P-3PA ORB-TRB AS ST BL TO PF TP
Earl,Acie C 44 13-23 14-19 0- 1 4-12 3 0 2 6 3 40

this is real? there was a game in which Acie Earl took 23 FGs and dropped a 40-12? The Bucks go wild

[Groundhog]

this is real? there was a game in which Acie Earl took 23 FGs and dropped a 40-12? The Bucks go wild

Indeed. December 4, 1996. A dark, dark day for mankind.

04/12/1996 NBA Box Score at BOS - basketballreference.com (http://www.basketballreference.com/teams/boxscore.htm?yr=1995&b=19960412&tm=BOS)

As a side-note (and the only reason I'm aware of this game), Acie Earl was like Shaq in his brief time downunder.

[AgustusM]

ALL of them

Statistics in sports generally suffer from low sample size, lack of context, and are considerably influenced by team, weather, coaching, style, etc, etc.

Some measures make sense - a basketball player who scores 25 points a game is probably better then one who scores 5 points and game. But to say a guy who averages 25.3 is better then a guy at 23.4 is absurd and doesn't take into account all of the above factors.

Likewise a RB who gains 1,500 yards is better than a guy at 500 - but again 1,500 compared to 1,300 doesn't tell you much.

I do understand we need something to help us understand what has happened in all of these games we never saw - but I believe as a sports culture we have gone way overboard in using numbers in an attempt to justify our inherent biases.

Now excuse me while I go check the PER numbers on the latest CP3 to the Warriors rumor while I pour over the box scores from one of my many text sims.

[Groundhog]

Drew Gooden proves your point. I remember when the Cavs got him I hadn't seen much of him at all and looked at the numbers he had put up, and I was excited. Now it still puzzles me at times how a guy can score and rebound like he often does and still hurt his team on both ends of the court at such a high and consistent rate.

Then he gets traded to the Bulls, their fans get excited, and so on and so on.

[Neon_Chaos]

PPG/48min

[stevew]

Defensive Rebounds are pretty questionable as a stat. It's pretty close to the quintessential "getting a stat for something you're supposed to do" stat. You want credit for that?

[TroyF]

A quality start is 6 IP 3 ER which equals a 4.50 era.

I've read a lot of analysis on this and the reality is rarely does a guy go 6 IP with 3 ER. As a rule, the teams and pitchers who have the most quality starts are the best in the game. And the guy with a 4.5 ERA doesn't have a lot of quality starts. When you look at quality starts stats, you can see certain things that jump out quickly.

For example, the Pirates finished with the 17th best ERA in the game last year. They had only 78 quality starts (average in baseball was 87) That tells you that those 78 quality starts were REALLY good. And when they didn't get that quality start, they were incredibly poor.

Texas finished 13th in MLB in ERA, but had 99 quality starts. The staff kept them in most games, which is all they really needed with that offense.

[CraigSca]

Any stat that doesn't provide context.

For instance, they'll sometimes post a stat like, "The Colts are 1 for 5 this year when going for it on 4th down." Is that good? Is that bad? Let me know what the median is for the league so I can at least put the thing in context.

Also, "loaded" statistics, like the inevitable, "The Stink Rats are 15-1 when Joe Runningback runs for 100 yards are more." This is useless because the statistic itself is influenced by what it's trying to measure.

[Danny]

Home Runs

[BYU 14]

Thread title says "sports".

Angry NASCAR fans in 3, 2 :)

[EagleFan]

Drunk NASCAR fans in 3..2.. :)

[Suburban Rhythm]

Drunk NASCAR fans in 3..2.. :)

What is the alternative?

[Young Drachma]

WAR.

[Autumn]

Indeed. December 4, 1996. A dark, dark day for mankind.

04/12/1996 NBA Box Score at BOS - basketballreference.com (http://www.basketballreference.com/teams/boxscore.htm?yr=1995&b=19960412&tm=BOS)

As a side-note (and the only reason I'm aware of this game), Acie Earl was like Shaq in his brief time downunder.

Ollie Miller, Acie Earl, Pervis Ellison ... wow, it's like the Convention of Pretty Crappy Centers I Still Watched and Cheered For From My Youth

[Groundhog]

Yah, looking at the lineups in that game, I mean... wow, what a dark era for the NBA.