Oh, they got outscored 8:23, lost the final, but somehow they crushed the Bruins in terms of ****ing CORSI? Well considering that this is a prospect top100 based on the Corsi system it should be best for the very new National Corsi League.
Hockey Prospectus Top 100 Prospects
- Thread starter BeardyCanuck03
- Start date
Oh, they got outscored 8:23, lost the final, but somehow they crushed the Bruins in terms of ****ing CORSI? Well considering that this is a prospect top100 based on the Corsi system it should be best for the very new National Corsi League.
Jason MacIsaac
Registered User
I guess absolutely lights out goaltending at one end vs piss poor goaltending at the other wouldn't cause in imbalance like that. Sorry if you fail to understand the stat, I'm sure someone will tutor you on the subject if you please.
So this is really your vision of the SC finals? Canucks crushed the pitiful Bruins with their puck possession skill but somehow got crushed on the scoreboard because Luongo is that much worse than Thomas? I must have watched some other games because it sure as hell was not that simple.
While the Corsi rating can be considered useful within context much like the +/- ratio does, it doesn't consider the quality of shots and many other factors. It is what it is - an overhyped new stat. I dont even know who should need tutoring on a formula as simple as Corsi, not much to understand there.
Benttheknee
Registered User
Does Corsi measure quality of shots? The Sens during their heyday would dominate the shots department and still fail to get far.
Corsi measures shot numbers, not quality, so a 2 man breakaway is rated the same as a blue line dump into the goalie hoping for a rebound that doesn't come.
Corsi can be quite useful, however it's not the sort of amazing be all end all stat that so many people try to make it out to be. It also does more to rate linemates than the individuals its assigned to. Wild fans lost count of the number of open nets that Miettinen missed after Koivu set him up perfectly last year. Koivu did the work to create the shot, but his bad linemate made it so he didn't get any stat credit for it (that goes for assists, +/-, and Corsi).
Well the Kabanov explanation is a whole other thing, for which I explain here:
http://www.hockeyprospectus.com/article.php?articleid=1036
In terms of Scheifele, here is my scouting report on him:
http://www.hockeyprospectus.com/article.php?articleid=1057
I don't think he has a whole lot of upside, but he certainly has above-average skills and I think being a fine second line forward is a reasonable projection for him. He's skilled, but not top-end, and I'm not that enthralled by his skating. I see him as a safe projection, not an upside one and that's more or less why I had him in the mid 20's before the draft.
I think by this time next year, I can certainly see him making the 75-100 on my list with an extra year of development under him.
Shot quality has been proven many times in the analytics world to be a mostly non-important factor over the long term. It exists, but multiple studies has shown it is not a repeatable skill of high significance.
http://www.arcticicehockey.com/2010/9/20/1696352/team-shooting-percentage-and-shot
http://www.hockeyprospectus.com/article.php?articleid=540
------
So what are these "other factors" you speak of, and where is your evidence?
Umm...you might want to...I don't know...read the things you post before describing what they say?
The first link sums up to "Shot quality clearly exists and has strong impact, but I haven't found a way to make it predictive."
The second link says "shot quality clearly exists, is strongly statistically significant, and clearly plays a role in scoring. It does not correlate to goals differential because of the strong inverse correlation between Corsi and shot quality. Small sample sizes make statistical analysis difficult. It strongly corrects for the large inaccuracies found in predictions utilizing Corsi."
In other words, the first link very very tenatively supports your claim, and the second link outright refutes it.
haha at everybody in this thread. Get 1 unconventional ranking system, and all of the amateur HF scouts come out in full force. I almost guarantee that everybody saying "OMG, player x over player y, 'nuff said, list is terrible", haven't even watched player X or player Y play, ever, and are basing their opinion on draft position, or other rankings.
this thread is like a microcosm of the first half of Moneyball.
I'll take Pronman's rankings against Hockey futures crappy rankings any day. Just because he uses different metrics to determine his rankings, and doesn't care where a guy was drafted, doesn't mean he is wrong.
The fact is, you can take any 2 players on the list, and Corey will actually have a very good reason for why 1 player was ranked ahead of the other. You just can't say that for most people's top 100 lists.
this thread is like a microcosm of the first half of Moneyball.
I'll take Pronman's rankings against Hockey futures crappy rankings any day. Just because he uses different metrics to determine his rankings, and doesn't care where a guy was drafted, doesn't mean he is wrong.
The fact is, you can take any 2 players on the list, and Corey will actually have a very good reason for why 1 player was ranked ahead of the other. You just can't say that for most people's top 100 lists.
The first link says,
"The amount of "shot quality" at the team level regresses roughly 50% in both cases, and the observed "shot quality" in one year or one half of the year is a very poor predictor of performance in the other half of the sample. The standard deviation of observed "shot quality" is on the order of 10 goals per season. But only 25% of this is skill, so the league-wide spread due to "shot quality" talent is on the order of six goals, or one win. (If we include data from the 1st and 2nd periods of close games, we end up with the same overall result.)
So it's not nothing. But it's less than the talent difference between Roberto Luongo and Michael Leighton (2.5-3 wins), for example. While "shot quality" may sometimes explain the role of luck in past performance, it has very little predictive value at the team level."
The second link says,
"The good news: the year-over-year correlation coefficient is 0.298, which means that shot quality will regress 70% year-over-year. The actual sustainability is a bit higher, for all the reasons mentioned above, but this is a good ballpark. This value is statistically significant, so sustainability does exist, as it should if my above analysis (that shot quality is not all luck but is at least in part due to skill) is true. "
I didn't say shot quality does not exist. I said it does not persist at a high degree of significance.
It's worth a handful of wins at most per season. The same author who said that above in the first link said,
"Together, Fenwick/Corsi and Luck account for around 3/4 of team winning percentage. What's the remainder? Goaltending talent - which Tom Awad estimates at about 5% - and special teams, along with a very small sliver that's due to shooting talent and the oft-mentioned "shot quality."
http://www.arcticicehockey.com/2010/11/25/1835371/retro-nhl-and-anger-at-corsi
Edmonton East
BUT the ADvaNCEd STatS...
- Nov 25, 2007
- 6,491
- 2,447
So anyone else find it interesting that out of the top 10, 8 are centers and 1 is a defender?
Ya, terrible ranking system.
Ya, terrible ranking system.
Avsfanftw
Registered User
Because people are simple minded and forget that he was being talked of as a top pick before his attitude problems.
Attitude problems =/= potential skill
pucka lucka
Registered User
Your list is quite awful. I am not sure why we're even discussing it. Grimaldi at least 70 spots higher than Scheifele? Bizarro world. Paul Postma just slightly behind Simon Despres?
TheBakester66
Registered User
Corey,
I just picked one quote of many to which i feel someone needs to respond. I was going to try to stay out of this thread as a fellow person willing to throw his lists up on HF boards to whatever criticism may come. And you can argue this guy over that guy until you're blue in the face. Opinions are opinions and I'll save my strength for when my list comes out.
My problem is really the model you've used to base your rankings. For a little background, I'm a software engineer from St. Louis, and my best friend is Dr. Stephen Shea, PHD in Math and Assistant Professor in the Math department at St. Anselm's College in New Hampshire.
We analyzed all of your sources. I gave these references more credit than Dr. Shea did, until he explained his reasoning of course.
A few conclusions that we drew.
A) The 38% chance of winning is luck is not actually backed up by anything in the reference you gave (a comment was actually made in the article to this effect), and what we did find is that Corsi can predict winning 60 something percent of the time, depending on what study and who you ask. In statistics, a 60 percent correlation is terrible at best, and would never be publishable by any reputable journal. More to my point, saying that Corsi has a correlation with Wins does not prove that other factors do not have a higher correlation to winning. I would bet that just by comparing the goaltender with the higher save percentage on the season as of the day of the game that you can predict win percentage at higher than 60 percent. Not that that's a strong model either, it just has nothing to do with Corsi and leads nicely to my next point.
B) You say that goaltending is easy to come by. Here's a point for you. The Flyers traded their two young franchise forwards to be able sign an over 30 year old goalie that is very hot and cold- dominant at times, miserable at times. And compared to the market, they significantly overpaid him to ensure his services. Your argument in this would probably be that the Flyers were significantly misguided in doing this. They weren't, but that's not my point. You also said that Goaltending can only account for 1 to 2 games per year. Dr. Shea and I are waiting for our paper on WORP (Wins Over Replacement Player) for goalies to be published in the Journal of Quantitative Analysis in Sports Statistics. With fairly simple and provable math, we were able to prove that the greatest number of Wins over replacement (depending on the exact method used) that a goalie was able to provide since 1982 is between 8 and 11 wins. Dominik Hasek had 4 seasons in this range. The worst goalies of all time were Peter Sidorkiwicz and Craig Billington for Ottawa in 1993 and 1994 when they had no one else, and lost their team roughly 5 to 9 games (again depending on method) a year because they were so much worse than average. The reasons the worst losses are lower than the most wins? Most teams will stop playing a goalie if he's ineffective. Ottawa literally had no other options than to play goalies with over 4 gaa for 50+ games. If you'd like the facts to back it up, I will be happy to provide our 13 page paper on the subject. So right there, from worst to best all time is a 13 to 20 game swing. Now obviously the teams play has a major impact as well, so it's certainly not apples to apples, but to say that a team could swing 10 games if they had the best goaltender in the NHL versus the 59th and 60th best goaltender in the NHL is probably not a stretch. That is 20 points. In almost all cases, 20 points is the difference between a team making or missing the playoffs.
C) Shot Quality is a very new area of study by which really only one guy (Ken Kryzwicki) has done extensive valid research, and only for the past 2 or 3 seasons. That is not enough data to conclusively state anything about Shot Quality. And saying that shot quality is not sustainable just sounds ridiculous. Coaches devote their life to sustaining shot quality. Maybe they all fail miserably, but I'd be inclined to guess that if I took Mr. Kryzwicki's work and analyzed this aspect of even his past two seasons of data, i would see team by team correlations on shot quality from game to game. What shot quality doesn't account for is the ability of the shooter relative to the average ability to score from each position on the ice, nor the events preceding the shot in question. A ridiculous one-timer at a wide open net (because the goalie is still on the other end of the net) is easier to score on then a shot from the same position with the goalie ready for the shot.
D) This notion that luck plays a part in winning games. How do you prove that luck wins games? Try and come up with a scenario under which luck would win a game. Replace "luck" with "margin of error in my statistical model" and everything that you say about luck makes more sense.
E) The notion that puck posession is by far the most important metric in rating prospects is flawed due to the conclusions drawn by the references you posted. We will acknowledge that having more shots than your opponent is most likely correlated with scoring more goals. And many have proven that scoring more goals than your opponent on average is correlated to winning more games. Those are both reasonable conclusions. What's not reasonable to conclude however, is that the ability to take more shots than your opponent is the most important factor in winning games. It easily could be argued that getting Joe Sakic 3 shots one-on-one per game against the goalie is more valuable than having Joe Sacco take 56 shots in a game. Really, what all of these corsi papers prove is that on average, teams do a fairly good job of balancing shot scoring ability, albeit through player abilities and style of play. This is most likely due to talent balancing rules like the Salary Cap and allowing the worst team to pick first in the draft. You can have Ovechkin on your team, but you have to burn 9 mil of your cap on him, whereas another team could instead go for two guys at 4.5 million. The combined scoring ability above average may just as well end up the same in both cases. So, we would argue that the only thing Corsi really proves is that teams emphasize scoring ability in drafting and signing players, and due to the cap, scoring ability is fairly balanced from team to team (so the cap is working). Thus scoring ability cancels out and makes shot totals look like the be all and end all and proving Corsi's worth. In reality, if you took a team made of all amazing puck possessors that could only shoot about as well as I do out there, your team would probably be horrible whether you outshot your opponent by 50 every game or not.
And finally: Professional franchises in the NHL may be behind the statistical times, and our own Billy Beane/ Bill James revolution may be right around the corner, but that doesn't mean that these guys are complete idiots either. You say goaltending is easy to come by, but in my opinion, every GM in the NHL would trade their top prospect for Jacob Markstrom if they had a major need in the net, like for example Phoenix or Tampa. Oliver Ekman-Larsson for Markstrom? Done. At worst, 75 percent of your list would be traded for Markstrom straight up. At best, 100 percent of your list would be traded for Markstrom straight up. Did you ask your internal NHL people about that?
Benttheknee
Registered User
Amazing stuff in this thread.
Since this is Corey Pronmans first kick at the 100 prospect can, I think that it is safe to say that we really do not know how accurate it is, but we will be able to measure how accurate is was in the future.
Has anybody done any type of analyis on who provides a better rank? Or has anybody gone through older drafts to rerank them to show who actually did draft well? That is what I think is interesting.
I have my own very rudimentry system, and am curious how it will fare without any special resources other than what is available to us couch GMs.
Since this is Corey Pronmans first kick at the 100 prospect can, I think that it is safe to say that we really do not know how accurate it is, but we will be able to measure how accurate is was in the future.
Has anybody done any type of analyis on who provides a better rank? Or has anybody gone through older drafts to rerank them to show who actually did draft well? That is what I think is interesting.
I have my own very rudimentry system, and am curious how it will fare without any special resources other than what is available to us couch GMs.
Uhh.. no. And not because goalies are easy to come by, but because they seem to have the worst correlation between good prospect and good NHLer. Remember when Lehtonen was the top HF prospect for like two years?
Thank you Corey for posting your list and for referencing some interesting statistics. I'd like to address the claim quoted above regarding "random chance" or what some replies have referred as "luck" determining the outcome of a game.
We often try to find a measure of how good a team is and then use it to determine on a given night if team A is better than team B. Suppose, according to our measure, Team A is better than Team B and Team A should beat Team B. Then the game is played and Team B wins. Because the better team did not win, we conclude "random chance" was involved. I don't accept this conclusion. It's probably more likely that the measure got it wrong.
I suspect, that as the season progresses, the ability of a team to win fluctuates. This could be due to injuries, coaches implementing a new strategy, starting goalie got a night off, etc... There are almost certainly factors current statistics can't account for.
If we use the measure that the team that wins the game is the better team that night. Than according to our measure, the better team wins 100% of the time. This is a bit silly, but it represents one end of a spectrum of possibilities. The possibility that random chance influences outcomes as much as 30-60% of the time, I contend, is the other end of the spectrum. The true answer probably lies somewhere in between.
Anyways, I am interested in any more references that use statistics to support random chance as a factor in determining the outcome of a game. Thanks.
Does anyone have the reference to the statistics to back up this claim? The article quoted doesn't seem to contain them. I fear that this statement may be a misinterpretation of correlation (but I can't be sure until I see the evidence).
Last edited by a moderator:
How many times have you seen Scheifele play in person ?
If you think at this time next year he will be in your 75 to 100 area , you will either be very surprised or , very wrong , or more likely both.
TheBakester66
Registered User
Goalie prospects can be difficult to project at 18 years old. When you're talking about 21 year olds that have already been awarded "goalie of the year" in the 3rd best league in the world, that is something entirely different. Especially now that we've had a year to watch him against north american competition. While i would argue that the success rate of goalies is equal to that of skaters, you would probably disagree. But what is a BUST for a goalie? Goalies seem to "BUST" in people's eyes because they are not bonafide NHL starters. Would anyone consider Cory Schneider a bust? All he's ever done is play well at every level he's been allowed to play at. Unfortunately for Cory, Vancouver is very happy with an elite and cheap backup to Luongo. Also makes for a great insurance policy if luongo gets hurt. No, what makes goalies deceptively prone to failure is lack of opportunity, not ability. Many goalies never go on to NHL success, but are extremely successful overseas. If you're not a top 30 pick, that still does not make you a bust. A lot of goalie progression is timing. You need a free backup spot at the AHL, a starting spot in the AHL, typically a year or two to backup at the NHL, and then you're finally given a starting gig. Or you can play overseas and take your chances when you come to America.
Another great example is Tim Thomas. Tim Thomas actually won an SM-Liiga (finland) championship with fellow american Brian Rafalski in 1998. After 4 years of NCAAs, Then 2 years in Finland, then a year in the AHL/IHL, then a year in Sweden, then a year in Finland, Then 2 years starting in the AHL, Then one more year back in Finland during the lockout, then another stint with Providence in the AHL, he finally got a midseason call-up in 2005-06, 11 years after he was drafted by the nordiques in 1994.
Does that mean he was a BUST? NO. He was phenomenal in Finland, Sweden, and the AHL. And even after having NHL success, he still nearly lost his job to a hotshot in Tuukka Rask. Which is another thing. Is Tuukka Rask a bust? Of course not, all he has done is play at an elite level. Still no starting spot.
Oh and as for Lehtonen, his problem was not performance/talent or opportunity, it was health. The fact that Lehtonen was able to come to the NHL and have success at such a young age shows how talented he really was. Of course you put no defense in front of him for 5 years and it's going to wear him down. He had a bit of Luongo in Florida syndrome in that respect. If Lehtonen was traded to Philadelphia or New Jersey, I think he could put up phenomenal numbers with the proper defense behind him. Certainly his stint in Dallas has proven that with some defense he's a lot better than with no defense.
Thanks for the detailed response. I appreciate the thoughts, even if I disagree with a lot of what you say.
To start off, I'm really, really not the guy to be the face of this conversation. There are much smarter guys in this area such as Tom Awad, Robert Vollman, Gabriel Desjardins, Vic Ferrari and J Likens.
The 38% was at the bottom of the article I linked to. That post wasn't a very through study but he got to the point and said in essence what he did, but Gabe has posted a ton of material on variance in NHL results. His e-mail is at the bottom of the site, I recommend if you want to get into a high-end debate about these topics, that you see him.
Yes 60% sucks, but most sports analysts will tell you 60% sucks in the real world, but it's amazing in the sports world, where so much of the game in affected by random chance.
The goaltending bit is correct on a seasonal basis. However Corsi is persistent which is why it is valuable. Goaltending performance fluctuates like crazy which does not give it a significant amount of long-term predictive value.
Goaltending in raw results are worth a ton. Tom Awad has published his Goals Versus Threshold (a hockey WAR measured in goals) results regularly for many seasons and goalies are always near the top. That's not why analysts are down on goalies though, it's because they can't repeat their performance (not to mention a flooded market of talent). And yes the Flyers made a horrendous move for what they did to get Bryzgalov.
I'm very familiar with the work of Alan Ryder and Ken. Ken and I actually used to talk quite a lot a few years ago when he was churning out a lot of material.
Ok then you need to prove it is sustainable and to what degree. Most researches that I have seen have shown it is not a highly persistent talent. If you have contradictory evidence, I would turn my opinion around.
Luck, random chance, variance, margin of error. All mean the same thing to me, just worded better.
Proof? You're saying I'm wrong, but why are you right?
I've referenced a lot of articles in this thread, my mistake if the one I did wasn't through, please DM me for what specifically you are looking for.
None. I've seen him about I'd say 10 times through video, but I never claim to be an independent scouting service. My work is heavily based on using my contacts throughout the hockey world to obtain scouting reports from. I combine knowledge from everyone I know, plus whatever I can watch of the player, and draw a conclusion. My scouting report of Scheifele is public. If you have an issue of my conclusion, state what part of his report is incorrect.
To start off, I'm really, really not the guy to be the face of this conversation. There are much smarter guys in this area such as Tom Awad, Robert Vollman, Gabriel Desjardins, Vic Ferrari and J Likens.
The 38% was at the bottom of the article I linked to. That post wasn't a very through study but he got to the point and said in essence what he did, but Gabe has posted a ton of material on variance in NHL results. His e-mail is at the bottom of the site, I recommend if you want to get into a high-end debate about these topics, that you see him.
Yes 60% sucks, but most sports analysts will tell you 60% sucks in the real world, but it's amazing in the sports world, where so much of the game in affected by random chance.
The goaltending bit is correct on a seasonal basis. However Corsi is persistent which is why it is valuable. Goaltending performance fluctuates like crazy which does not give it a significant amount of long-term predictive value.
Goaltending in raw results are worth a ton. Tom Awad has published his Goals Versus Threshold (a hockey WAR measured in goals) results regularly for many seasons and goalies are always near the top. That's not why analysts are down on goalies though, it's because they can't repeat their performance (not to mention a flooded market of talent). And yes the Flyers made a horrendous move for what they did to get Bryzgalov.
I'm very familiar with the work of Alan Ryder and Ken. Ken and I actually used to talk quite a lot a few years ago when he was churning out a lot of material.
Ok then you need to prove it is sustainable and to what degree. Most researches that I have seen have shown it is not a highly persistent talent. If you have contradictory evidence, I would turn my opinion around.
Luck, random chance, variance, margin of error. All mean the same thing to me, just worded better.
Proof? You're saying I'm wrong, but why are you right?
I've referenced a lot of articles in this thread, my mistake if the one I did wasn't through, please DM me for what specifically you are looking for.
None. I've seen him about I'd say 10 times through video, but I never claim to be an independent scouting service. My work is heavily based on using my contacts throughout the hockey world to obtain scouting reports from. I combine knowledge from everyone I know, plus whatever I can watch of the player, and draw a conclusion. My scouting report of Scheifele is public. If you have an issue of my conclusion, state what part of his report is incorrect.
TheBakester66
Registered User
The work of Tom Awad and others at hockey prospectus is a glorified educated guess at the goals above replacement for players. They cannot prove their formula because they pull numbers out of the sky to account for the value of players at one position to another, mainly because the results look okay. Well looking okay and being right are two different things. I'm still baffled by GvT's overuse in the mainstream, but most people don't bother to research it's origins. Now, bashing something is not very fair unless you're willing to provide an alternative solution, and to this point it seems fairly difficult to attempt to do what HP has done in any sort of provable way. The data just simply isn't there, or is only there for the past year or two. Doesn't mean we should be relying on GvT for anything. It just means that their educated guess is better than anything else that's been done at the moment. Which is fine for an individual's own interest, unusable in the published world. I know this isn't your fight, this is more for your own education.
In the future, you're really going to want to stop mentioning that so much of the game is affected by random chance. What's happening is that our statistical models are not strong enough, or we don't have enough data to say for certain who should win each game and why. Look at it this way. Give me one scenario in a real world game under which random chance decides who wins. I would argue that whoever plays better that night wins. Having more talent, or playing better prior to that game does not guarantee that you will play better that next night. There is nothing random about winning a game. Random chance would decide the winner if after a tie in regulation they decided to flip a coin to see who won the shootout instead of actually doing the shootout. Nothing in a hockey game is random.
My real beef with this statement is that looking at raw goaltending results is not the whole picture. See: Osgood, Chris or Lalime, Patrick. Some goaltenders are helped or hurt tremendously by the defense in front of them. Team systems rely more on the goaltender or less on the goaltender depending on the teams abilities and the goaltenders abilities. However, even defenses on the same team change from season to season. I would argue (though I have absolutely 0 evidence to back this up) that it is all of the changes going on around goalies that make raw goaltending statistics volatile. If the ability of goaltenders could not be relied on with any consistency, goaltenders would never sign long-term deals, and my guess is that RFAs would be used for the most part to help with the salary cap. I'd like to see a reference to an article that proves that goaltending ability is volatile, or not sustainable, where they attempted to factor out the surrounding results. I mean my guess is that if you had any NHL team play some crappy team for the entire season in 1985. And then replayed the season, but instead you went against the Edmonton Oilers every game for the season, your raw goaltending stats would be very different, but it wouldn't mean that your goalie was any worse, just meant that you went against an all-time offense. That's an unrealistic extreme example, but the point is simple.
I'll get to work on it right away. I'm joking, but it would be rather fun to do. So maybe I will.
Proof: If you can't score, you can't win, whether you dominate puck possession or not.
I'm not saying puck possession isn't important, it is. But some players can dangle to create space, and some players hit guys to create space. Others are exceptional passers that create space. Some are speedsters. Some are good at taking the puck away, some are good at never giving it up. And in the end, without guys that can consistently score, you can't win games. Ask the 05-06 Blues how easy it was to win games when Mike Sillinger (who played 3rd or 4th line checking btw) led the team in goals with 22.
I would say almost every goal scored has an element of ranDom chance in it. These are the best players in the world, but in the heat of the moment, in a game, when they take a shot, they are really aiming for 'an area' of the net. This area is hard to quantify, but it would be a smaller, or more precise area for the better shooters, and a larger, less precise area for the worse shooters. Alternatively, you have a goaltender who is covering his angles as well as possible. The shooter takes his shot, and the goalie attempts to make his save. The strength of the goalie is going to determine how small the margin of error will be for the shooter. For very good goalie, there may only be a 1-2 puck-sized hole in which he can hit for the puck to go in, and he may only hit that shot 1 out of 10 times. So, for any given goal, there is going to be a tremendous amount of luck involved. Now this is obviously a generalization, as every goal and every shot is a unique situation, and this is obviously very hard or impossible to quantify. But luck definitely plays a role in almost every area of the game; probably a very significant role.
For a specific example, look at game 7 of the first round series in 2011, Vancouver vs. Chicago. On the powerplay in overtime, Patrick Sharp recieved a good pass from behind the net and attempted to 1 time it into the top corner. Due to the speed of the play, there was a lot of open net to shoot at. Given it was a one timer, the margin of error was a lot higher. Patrick sharp is one of the best shooters in the league. Luongo did a good job of getting over, but I bet if we repeat that play 10 times, Sharp puts it in the top corner 8 out of 10 times. There was certainly an element of skill that went into the save, and the missed shot, but you could argue that it was very unlucky for that shot to not go in the net. And the Blackhawks were very unlucky to not advance to the second round of the playoffs. Now Canucks fans would probably point to a variety of instances in which the Hawks were lucky to even get to the 7th game, and they would probably be right. Arguing that luck does not play a significant role in hockey is pretty asinine.
We have different definitions of random chance. For me, an event in hockey is random if it is not in any way the result of the skills of the players involved. I would describe random chance as what happens when you roll the dice in a board game. You described a situation where the skill sets of both Sharp and Luongo were put to the test. In this case, Luongo stopped the shot. That's a direct consequence of the type of shot Sharp put on net and Luongo's ability to stop it. You may be correct that if the same situation came up again, the outcome may have been different. To me, that does not imply random chance. In baseball, a player will only hit a homerun every so many at bats, but you don't hit a 95 mph fastball 400 ft by chance. You have to have a certain set of skills. The same holds for shooting a puck on an NHL goaltender. It's not rolling dice.
If we say that, for example, 20% of games are determined by random chance, we're saying that the teams that won those games would have won regardless of their efforts on the ice. Clearly, that's never the case.
Last edited:
This is a very interesting and important debate. If a certain percentage of hockey games are determined by random chance, then there is a theoretical cap (which is less than 100%) to the accuracy of any predictive model. If there is no random chance, then the theoretical cap is 100%. Is what we perceive as unpredictable by current modeling methods truly the result of this "luck," a force beyond human control, or with more knowledge, with a better understanding of the variables at play, could we improve our predictions?
