Simple Goals vs Assists frequency graph

[andrjusha]

Looking through nhl records for goals and assists per season
NHL Records
plotted frequency graph, eg 90+ goals achieved 1 time, 90+ assists achieved 27 times, etc

Interesting to analyze, eg 100 frequency is for 74A and 53G which is 1.39 ratio
1000 frequency is for 50A, 36G which is 1.39 ratio
20 frequency for 93A, 67G, again 1.39

Does it mean that 1 Goal is equal to 1.39 Assists, naively thinking?
Scoring rates are different for different eras but what do these bulk data tell us about ballpark value of goals vs assists?

Click to read more...

 

[Czech Your Math]

I doubt there's anything really indicative of the ratio of G to A at each frequency being lower than goals to assists in every day play. The better goal scorers tend to score substantially fewer goals than the better playmakers get assists. The lower totals are more susceptible to random variation, since the sample size of that stat is lower (i.e., it's more likely a "40 goal scorer" scores 53 one season than it is a "55 assister" gets 74 assists one season).

 

[andrjusha]

average goals/assist ratio in NHL is 1.6-1.7 range, that 1.39 number is pretty stable for different frequencies as well,
wonder what are other G/A weights people use and reasons

 

[Czech Your Math]

average goals/assist ratio in NHL is 1.6-1.7 range, that 1.39 number is pretty stable for different frequencies as well,
wonder what are other G/A weights people use and reasons

The tails are just going to be fatter when the counts are lower. If you give a class a 25 question quiz, it's more likely that a student can get lucky with a few correct guesses and get a higher grade, than if they were give a 35 question quiz. I think that's why the ratio seems to be pretty stable, albeit at a lower value.

As far as weights for G/A, there's probably two main modes of thinking:

A) Weight them equally, as they each produce a goal in the end.
B) Weight them inverse to the ratio at which they occur (scarcer is more valuable)

I agree mostly with A, although I see the logic of B too.

 

[Zuluss]

I found out that many people implicitly agree than 1:1.5 ratio is fair as far as valuing goals vs. assists, and that lands in between the 1:1.4 ratio from the OP and 1:1.7 based on the frequency of goals and assists recorded.

The reason why I say "implicitly agree" is that people would frame it as "20 extra goals would cover a 10-point gap", as in "50g+40a is as good of a stat line as 30g+70a" - or 50*1.5+40=115=30*1.5+70
Then again, it probably does not matter all that much if you use 1.4 or 1.7 in this comparison: 1.4*50+40=110, 1.4*30+70=112; 1.7*50+40=125, 30*1.7+70=121. 50g+40a is subjectively close to 30g+70a anyway.

 

[txpd]

The last 82 game schedule there were 2 50 goal scorers and 32 50 assist scorers. I'd say that says it well. This is the argument we see when we try to evaluate how good Ovechkin really is. You reduce him nicely by making a goal and an assist the same value

 

[Czech Your Math]

The last 82 game schedule there were 2 50 goal scorers and 32 50 assist scorers. I'd say that says it well. This is the argument we see when we try to evaluate how good Ovechkin really is. You reduce him nicely by making a goal and an assist the same value

One could look at it that way, but aren't you devaluing more balanced/versatile scorers or even playmakers in general by using a lower ratio? Let's say a player is balanced, scoring a good number more assists than goals, but plenty of goals too. Would his team be better off if he scored more goals, but fewer overall points? I don't think so, at least not generally. Uniqueness in this case doesn'tn necessarily translate to value. There are at least some players that likely could have scored more goals (perhaps enough to win goal-scoring title), but decreased their total points in the process. That doesn't seem more valuable to me. Could Ovechkin have decreased his goal value to score more totals points? If so, why didn't he? Wouldn't that be more valuable than just scoring more goals by itself?

 

[txpd]

One could look at it that way, but aren't you devaluing more balanced/versatile scorers or even playmakers in general by using a lower ratio? Let's say a player is balanced, scoring a good number more assists than goals, but plenty of goals too. Would his team be better off if he scored more goals, but fewer overall points? I don't think so, at least not generally. Uniqueness in this case doesn'tn necessarily translate to value. There are at least some players that likely could have scored more goals (perhaps enough to win goal-scoring title), but decreased their total points in the process. That doesn't seem more valuable to me. Could Ovechkin have decreased his goal value to score more totals points? If so, why didn't he? Wouldn't that be more valuable than just scoring more goals by itself?

I am not devaluing anything. Its simple. 2 50 goal scorers vs 32 50 assists scorers makes perfect sense when 2 assists are awarded on most goals. I dont object all to 2 assists being needed, but the numbers are pretty clear.

This may be news to you, but Ovechkin is the best goal scorer in the league. Arguably of all time. The offense is set up for him to shoot. The power play set up is not set up for any player on the ice to shoot off his pass. Thats not because he cant pass. The power play is set up doesnt allow for players to either score his rebounds or deflect his shots. Thats not because he doesnt record shots on goal.

Early in Ovechkin's career the Caps were losing by a goal late in a game 7. Ov had a clear scoring chance and he passed the puck to a wide open Sergei Fedorov for a lay up. The pass was on target but Fedorov missed. The question after they lost was why did he pass up a chance to tie the game. That is his job.

Ovechkin has the best one timer in the league. Nicklas Backstrom cant shoot one.

 

[Czech Your Math]

I am not devaluing anything. Its simple. 2 50 goal scorers vs 32 50 assists scorers makes perfect sense when 2 assists are awarded on most goals. I dont object all to 2 assists being needed, but the numbers are pretty clear.

This may be news to you, but Ovechkin is the best goal scorer in the league. Arguably of all time. The offense is set up for him to shoot. The power play set up is not set up for any player on the ice to shoot off his pass. Thats not because he cant pass. The power play is set up doesnt allow for players to either score his rebounds or deflect his shots. Thats not because he doesnt record shots on goal.

I just don't understand how a player scoring more goals, but fewer total points can be said to be more valuable than one with more points, but fewer goals. They could be, but I would guess they generally aren't. If they system is geared towards Ovi scoring goals and everyone knows that, I'm not sure it's the best system, since predictability is easier to defend. Again, it could be, but I would guess it's usually not.

Early in Ovechkin's career the Caps were losing by a goal late in a game 7. Ov had a clear scoring chance and he passed the puck to a wide open Sergei Fedorov for a lay up. The pass was on target but Fedorov missed. The question after they lost was why did he pass up a chance to tie the game. That is his job.

Ovechkin has the best one timer in the league. Nicklas Backstrom cant shoot one.

If he made the play that he felt had the best chance of them winning the game, then it was logically correct. If he made an incorrect assumption, then it was still good logic, but perhaps a bad decision. If it was simply bad results, then that happens. Ovi could have missed the scoring chance too, but that doesn't really affect whether the decision was right or wrong.

 

[Name Nameless]

I think one of the things I love the most with hockey, is by counting assists, people realize this is a team-sport. With football, or soccer as you call it, all the talk is about the goalscorers. Well, I guess some people wish they were football (or, sorry, soccer) - fans then.

 

[Czech Your Math]

I think one of the things I love the most with hockey, is by counting assists, people realize this is a team-sport. With football, or soccer as you call it, all the talk is about the goalscorers. Well, I guess some people wish they were football (or, sorry, soccer) - fans then.

Yes and it's tough to generalize. Sometimes at least one of those getting the assist really made the play that resulted in the goal, while other times they did nothing above standard or even routine, and were just fortunate the goal-scorer or another player made a good play. The proportion of those types of plays can vary dramatically from player to player.

 

[Name Nameless]

Yes and it's tough to generalize. Sometimes at least one of those getting the assist really made the play that resulted in the goal, while other times they did nothing above standard or even routine, and were just fortunate the goal-scorer or another player made a good play. The proportion of those types of plays can vary dramatically from player to player.

What you ignore in your math when you want to devalue the goal the assister gets an assist on, is most players who gives great passes, also gives great passes where the so-called goal-scorers manages to blunder given goals.

 

[Czech Your Math]

What you ignore in your math when you want to devalue the goal the assister gets an assist on, is most players who gives great passes, also gives great passes where the so-called goal-scorers manages to blunder given goals.

I'm not the one suggesting devaluing assists. As I said before, I generally believe assists are as valuable as goals or at least close.

 

[Lexus]

I think this stat is pretty wild to be honest...

Points by Peter Bondra (GP/Goals/Assists/Points)

1994–95

Washington Capitals

NHL

47

34

9

43

[TBODY] [/TBODY]

Funny thing is that he had more goals than assists during his time in the NHL

NHL totals	1081	503	389	892	749

[TBODY] [/TBODY]

 

[Dingo]

they should be separated as a stats. they describe a different play and typically very different players receive many of each.

 

[Dingo]

I am not devaluing anything. Its simple. 2 50 goal scorers vs 32 50 assists scorers makes perfect sense when 2 assists are awarded on most goals. I dont object all to 2 assists being needed, but the numbers are pretty clear.

This may be news to you, but Ovechkin is the best goal scorer in the league. Arguably of all time. The offense is set up for him to shoot. The power play set up is not set up for any player on the ice to shoot off his pass. Thats not because he cant pass. The power play is set up doesnt allow for players to either score his rebounds or deflect his shots. Thats not because he doesnt record shots on goal.

Early in Ovechkin's career the Caps were losing by a goal late in a game 7. Ov had a clear scoring chance and he passed the puck to a wide open Sergei Fedorov for a lay up. The pass was on target but Fedorov missed. The question after they lost was why did he pass up a chance to tie the game. That is his job.

Ovechkin has the best one timer in the league. Nicklas Backstrom cant shoot one.

i remember that pass to Fedorov well. seemed like he missed it because everyone expected a shot from Ovie at that point. Thats what made the slap pass such a brilliant decision... only it also fooled Fedorov.

 

[Zuluss]

I just don't understand how a player scoring more goals, but fewer total points can be said to be more valuable than one with more points, but fewer goals. They could be, but I would guess they generally aren't.

Some players collect secondary assists from carrying the puck into the zone and/or playing on the half-wall on PP. Other players are designated triggermen on PP who never touch the puck and have to produce a great shot pretty much every time they collect a PP point. Generally the latter are more impactful than the former if points are the same.

 

[Czech Your Math]

Some players collect secondary assists from carrying the puck into the zone and/or playing on the half-wall on PP. Other players are designated triggermen on PP who never touch the puck and have to produce a great shot pretty much every time they collect a PP point. Generally the latter are more impactful than the former if points are the same.

Plausible, but then if a player is almost never carrying or touching the puck, other than to shoot... there's usually a reason for that.

 

[Zuluss]

Plausible, but then if a player is almost never carrying or touching the puck, other than to shoot... there's usually a reason for that.

This reason is called comparative advantage: if player A is a better goalscorer than playmaker and the opposite is true of player B, B will be doing the passing and A will be doing the shooting - even if player A is overall so much better than Player B that A is also a better playmaker than B. Still, you cannot have Player A score from his own passes, so Player B will be the playmaker in the pair.

 

[The Panther]

At the risk of stating the obvious, no hockey fan on earth thinks a 50-assist season is the equivalent of a 50-goal season. So, that was sort of a silly comparison.

There is an arbitrariness to how assists are awarded in the NHL (two per goal, depending on teammate 'touches' or not) and how points are accumulated (assist of any sort equal to a goal). But no matter how you do it, there is going to be some arbitrariness. Naturally, there are extreme examples of high-scoring forwards who had huge-goal/very-low assist seasons and of high-scoring forwards with huge-assist/very-low goal seasons, but these rarely result in Art Ross winners either way. Most statistical evidence I've seen says that deleting 'secondary assists' or devaluing them in some way would have fairly little effect on historical scoring leaders.

The really bizarre point of NHL history is how there was no 'Goals trophy' (now the Maurice Richard trophy) until 1999. The NHL must be a unique league around the world for that!! It took 72 years to get that award! Just weird. How the hell do you have an award for a "Defensive forward" (whatever that is) and no award for the guy who scores the most goals...?

I think, if I had to change the NHL point-scoring system (and I don't really want to), I would change it in one of these two ways:

1) Award only "1/2" a point for a secondary assist. This would, of course, screw some players who make the key play on certain goals, but it would all come out in the wash. The main problem with this system is we'd end up with half the leading scorers having 90.5 points or something, which just looks stupid. (A bunch of hockey things already appear nonsensical and arbitrary to first-time viewers of the sport, and this kind of thing would just add fuel to the fire.)

Or (and I think this would actually be better):

2) Keep it just as it is, with the max. of two assists per goal, BUT the off-ice officials should determine whether or not a secondary assist is "deserved" or not. There would of course have to be an objective way to do this, so I suggest that if the "secondary" passer either (a) makes and completes his pass on his own side of the red line, or (b) doesn't "direct" the puck with his stick in any discernible way (e.g., if he shoots from the point and it bounces off two teammates and goes in), then no secondary assist is awarded.

The latter system would only cut back on secondary assists by a small margin, but I think it would be useful just to weed out the really silly point awarding.

But, basically, I'm okay with it the way it is.

 

[Hockey Stathead]

I think this stat is pretty wild to be honest...

Points by Peter Bondra (GP/Goals/Assists/Points)

1994–95

Washington Capitals

NHL

47

34

9

43

[TBODY] [/TBODY]
Funny thing is that he had more goals than assists during his time in the NHL

NHL totals	1081	503	389	892	749

[TBODY] [/TBODY]

Bondra just misses making this list with 25 more goals than assists in 1994-1995.