Frequency of Hat Tricks

[PrimumHockeyist]

Hi all.

Here out west, I've come to expect that visiting teams will set big records when they're on the verge. It's what we do here in Vancouver.

And so tonight, when Stamkos entered town with 499 goals it was not unexpected that he scored. Now, as we enter the third period Stamkos is entertaining a hat trick, having scored two goals this far. The announcer, John Garret, just said that he would be the 8th such player to score a hat trick when he scored his 500th goal.

So, until tonight there had been 46 500 goal scorers.
If 7 scored hat tricks on the night of their 500th goal, the frequency is one hat trick every 6.75 games on this milestone occasion.

Can anyone tell me if that anything close to the normal frequency of hat tricks?

A further consideration would be the average age of players when they score 500. It's probably too much of a pain to figure out, but I wonder how often players score hat tricks at that age.

Click to read more...

 

[PrimumHockeyist]

Can anyone tell me if that anything close to the normal frequency of hat tricks?

Here's a sketch:

NHL Hat Tricks History & Fun Facts December 12, 2022 by Matthew Zator

"A hat trick is one of the NHL’s rarest achievements with only 102 during the 2021-22 season."

If the writer is talking about regular season stats, there would be 1312 regular season games to consider, and hat tricks would be scored about once every 13 games. Or, to be more exact, one might say that two teams combine to produce one hat trick about every 12.9 games = 1312/102.

But the question I asked was about players, How often do 'players' score hat tricks on the same night that they score their 500th goal. With Stamkos' trio last night, the frequency now 8 hat tricks over 47 games, or once every 5.8 games. When a player scores his 500th goal, there is a better than 1 in 6 chance that he will score a hat trick. This seems remarkable, given the previous comparison, since single players are basically doubling the rate set by two teams in 20-21....

This makes me wonder what the average is among forwards. The same article mentions "a 15-season veteran like Claude Giroux, who has over 300 goals, only has one regular-season hat trick." Claude's rate of frequency is about 1 in 300, He scores hat tricks in 0.03 percent of his games. Maybe this is a bit low among, let us say, Top 6 forwards.

And of course the writer mentions Gretzky, who may set the highest hat trick frequency standard, which does in volume: " Wayne Gretzky had 50 regular season hat tricks in his career, or one in 3.4 percent of his games." Here we can say that this is based on Gretzky's regular season and playoff NHL games, 1695, since the math checks out (1695/50).

When the NHL's 500th goal hat trick frequency, of 8 in 47 games, is projected to 1695 games, 288 guys can be expected score hat tricks on the same night that they score their 500th goal.

-------- One further thing--.

In 81-82 and 83-84, The Great One netted 10 hat tricks over the course of two 80-game regular seasons. What may be more amazing, is that Gretzky's truly astonishing performances are significantly less than what can be expected on the night when players score their 500th goals, based on a solid sample of NHL data.

When similarly projected, the current 8-hat-tricks-every-47 games rate translates to 13.6 hat tricks per 80-game season! This is 36 percent more productive than Gretzky ever was in his finest 'hat trick' seasons.

 

[Hockey Outsider]

Using basic probability theory (Poisson distribution), a player whose "true" talent level is 40 goals per 82 games, would have (roughly) a 1 in 74 chance of getting three (or more) goals in any given game.

So, if we simplify things and assume ALL players approaching 500 goals are scoring at a 40 goal pace, we only would have expected one player to get a hat trick in the game where they reach 500.

Some players barely limped over the finish line (Pierre Turgeon, Peter Bondra, Jeremy Roenick, Lanny McDonald - and I'm sure there are others). Their probability would be much lower (a fraction of 1%).

On the other hand, some players were at/near their prime when they reached 500 (Gretzky, Lemieux, Ovechkin, Bossy). But even someone like Lemieux (who was very nearly at 1 goal per game in the year he reached 500) only has about a 1/12 probability. And, obviously, only a few players who reached 500 goals were scoring anywhere close to that level at the time.

This shows me that, if eight players scored a hat trick in the game where they reached 500 goals, it's much more than we'd expect from basic probability theory. It could be pure randomness, or it could be because the player is particularly motivated when they're approaching a major milestone.

 

[morehockeystats]

Modern day hat-tricks often include an EN goal. In the past era they were much fewer. The OT goals are also somewhat easier to come by these days.

 

[PrimumHockeyist]

This shows me that, if eight players scored a hat trick in the game where they reached 500 goals, it's much more than we'd expect from basic probability theory. It could be pure randomness, or it could be because the player is particularly motivated when they're approaching a major milestone.

It sure is a mystery. the thing that I find most striking, given the sample size, which currently stands at 47 games, is how new members of the 500 club greatly outperform Gretzky's finest 'hat trick' seasons. For once, the separation seems to be going the other way. No surprise that his 500th fits the pattern.

 

[PrimumHockeyist]

Modern day hat-tricks often include an EN goal. In the past era they were much fewer. The OT goals are also somewhat easier to come by these days.

Good point, which takes us to the start of the 1983-84 season.

At a glance, I would guess that the earlier era brings the current 8 hat tricks out of 47 games rate down. That is, if there had been OT from the beginning, the number of hat tricks scored over the same 47 games would be higher than 8.

 

[Chili]

Modern day hat-tricks often include an EN goal. In the past era they were much fewer. The OT goals are also somewhat easier to come by these days.

Would be interesting to compare empty net goals per game average over the years. And believe it's hard to compare eras with rule changes like ot and removing the red line..

 

[pnep]

Would be interesting to compare empty net goals per game average over the years. And believe it's hard to compare eras with rule changes like ot and removing the red line..

League scoring is over 3 goals per game (per team) for the first time since 1995-96

Scoring has seen a drastic uptick over the past 5 years, and is now on par with that of 1995-96. Obviously it could level out, but the trend seems to be that scoring is increasing year after year. Do you think this will continue and we will eventually see something close to the 80s? Or will...

forums.hfboards.com

 

[Chili]

League scoring is over 3 goals per game (per team) for the first time since 1995-96

Scoring has seen a drastic uptick over the past 5 years, and is now on par with that of 1995-96. Obviously it could level out, but the trend seems to be that scoring is increasing year after year. Do you think this will continue and we will eventually see something close to the 80s? Or will...

forums.hfboards.com

Nice work, thanks!

 

[PrimumHockeyist]

Would be interesting to compare empty net goals per game average over the years. And believe it's hard to compare eras with rule changes like ot and removing the red line..

I agree. I would say it's pretty impossible. Case in point, a few months ago I wrote an essay explaining that Paul Henderson's exquisitely timeed Summit Series winning goal versus the Soviets in 1972 was - as a third straight game winning goal - something that he could be expected to do one game every 69 NHL seasons. That was based on his NHL data, and a solid sample size that occurred during a time when the best players were ending their careers with game winning goals just under once every ten games.

By that measure, the most elite players could be expected to score a third straight game winner just once every 12 seasons. That rate seems to have held up fairly well since the introduction of OT in 83-84, but it must have been lowered. Now, with goals counting in 3 on 3, this measure will have to change significantly.