Minnesota Golden Gopher Season 2018-2019: Fire Motzko

[D2D]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

IMO, the decline is due to moving to the Big 7.

It's not nearly as simple as that. Gophers lost to Holy Cross in the NCAA 1st Round way back in 2006, well before their supposed decline due to their moving to the "Big 7". Upsets across all conferences are happening with increasing regularity (just ask SCSU fans).

And three of the four teams that made it to last year's Frozen Four were B1G teams. After a rough start, it's become a very good conference.

 

[Koho]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Well there goes my bracket....

 

[Bonin21]

Well there goes my bracket....

Shame...

 

[D2D]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Well there goes my bracket....

Shame...

I would be shocked if anyone honestly gets them all right this year.

 

[The Rube]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

It's not nearly as simple as that. Gophers lost to Holy Cross in the NCAA 1st Round way back in 2006, well before their supposed decline due to their moving to the "Big 7". Upsets across all conferences are happening with increasing regularity (just ask SCSU fans).

And three of the four teams that made it to last year's Frozen Four were B1G teams. After a rough start, it's become a very good conference.

I don't recall when the NHL Rookie Max Salary came along, but that had significant impact, IMO. Why play another year in college if some team offered you the max? Sure, it might be because you don't wanna switch between the minors and NHL, but then again, do you want to risk an injury or rescinded offer? Took some teams (like MN) to adjust to that.

 

[Koho]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

I don't recall when the NHL Rookie Max Salary came along, but that had significant impact, IMO. Why play another year in college if some team offered you the max? Sure, it might be because you don't wanna switch between the minors and NHL, but then again, do you want to risk an injury or rescinded offer? Took some teams (like MN) to adjust to that.

I think an equal contributor, maybe bigger, has been the trend of the BlueBloods from the 90's and earlier recruiting 18 yr olds to now committing to 16 yr olds and younger. Way more misses for those teams on recruiting and when there is a kid who develops late, these teams have committed all their scholies and so the late bloomer is available to programs that didn't get such talented players regularly in the past. And when the Blueblood team do hit the jackpot with a player who takes off in skill, because of the NHL CBA, they don't stick around more than a year or two. Hence, the streak of one seeds losing in the first round since 2006 (think that is correct) and number of 4 seeds who've won it all.

 

[SteveO]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

I think an equal contributor, maybe bigger, has been the trend of the BlueBloods from the 90's and earlier recruiting 18 yr olds to now committing to 16 yr olds and younger. Way more misses for those teams on recruiting and when there is a kid who develops late, these teams have committed all their scholies and so the late bloomer is available to programs that didn't get such talented players regularly in the past. And when the Blueblood team do hit the jackpot with a player who takes off in skill, because of the NHL CBA, they don't stick around more than a year or two. Hence, the streak of one seeds losing in the first round since 2006 (think that is correct) and number of 4 seeds who've won it all.

IMO you're right on Koho! The 2006 CBA that resulted from the 2004-05 NHL lockout brought sweeping changes to hockey’s entire landscape. In addition to new on-ice rules, there were many changes to how players’ contracts could be structured, when players could sign and how much they could be paid on entry-level deals within the new collective bargaining agreement.

As a result, the NHL got younger in a hurry, which has had a lasting impact on college hockey. IMO I think we are now seeing the long term effects of top college recruiting strategies in regards to early departures that have not changed much since then. The flight to the NHL escalated after the 2006 CBA. The funny thing was Bettman tried to sell it to college hockey coaches as a good thing that would benefit recruiting, but it did just the opposite.

As of the 2017-18 season, 311 players (32%) who have seen at least one regular season NHL game were college players.

<img src="https://cdn3.sportngin.com/attachments/photo/1013/4400/CollegeNHL2018a.jpg" class="irc_mi" style="margin-top: 10px;" height="400" width="600">

19% of the 311 are from Minnesota.

<img src="https://cdn3.sportngin.com/attachments/photo/1013/4406/CollegeNHL2018b.jpg" class="irc_mi" style="margin-top: 10px;" height="400" width="600">

 

[trixR4kids]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Hence, the streak of one seeds losing in the first round since 2006 (think that is correct) and number of 4 seeds who've won it all.

Sorry but the one and done flukey tournament results don't really prove anything other than the fact that a single game of hockey can have a ton of variance in outcomes. If you want to argue that the blue bloods are missing the tournament more often or backing your assertion with something that has a significant sample size then that'd be reasonable (and I haven't looked into it but that argument might actually have some legs to it).

If your larger point is that there's more parity in college hockey, then I agree. But that's probably the worst way to measure/test that hypothesis.

 

[St. Clown]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

I would be shocked if anyone honestly gets them all right this year.

After Saturday, I read somewhere that only 0.3% of brackets submitted online were still 100% correct.

 

[dxmnkd316]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Sorry but the one and done flukey tournament results don't really prove anything other than the fact that a single game of hockey can have a ton of variance in outcomes. If you want to argue that the blue bloods are missing the tournament more often or backing your assertion with something that has a significant sample size then that'd be reasonable (and I haven't looked into it but that argument might actually have some legs to it).

If your larger point is that there's more parity in college hockey, then I agree. But that's probably the worst way to measure/test that hypothesis.

This is an excellent post. You and I have always agreed on this and the numbers.

 

[SteveO]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Sorry but the one and done flukey tournament results don't really prove anything other than the fact that a single game of hockey can have a ton of variance in outcomes. If you want to argue that the blue bloods are missing the tournament more often or backing your assertion with something that has a significant sample size then that'd be reasonable (and I haven't looked into it but that argument might actually have some legs to it).

If your larger point is that there's more parity in college hockey, then I agree. But that's probably the worst way to measure/test that hypothesis.

A small sample size and error variance can be corrected by measuring the variance of the prediction error while supplementing it with scatterplot graphics. In this case, expanding hypothesis testing should include the regular season, as well as regional game and FF correlations in the data analysis Koho mentioned.

"worst way to measure/test that hypothesis"

 

[trixR4kids]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

So we agree, you'd probably want to expand your analysis to include the regular season (thus increasing the sample size) to test the hypothesis?

 

[SteveO]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

So we agree, you'd probably want to expand your analysis to include the regular season (thus increasing the sample size) to test the hypothesis?

For higher internal validity, and optimal predictors in the sense of reduced prediction error variance using linear regression yeah it's a better choice. If I understand Koho's statement correctly, he's including regional games as well. So that would increase statistical power a bit too.

 

[trixR4kids]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Even if you include regional games you’re still talking about a one and done scenario where even in the most lopsided matchups the underdog probably has a 30% chance of winning. The results of a one and done won’t tell you nearly as much as a sample that’s 40x greater.

I don’t think AIC winning a single game tells us as much about the parity of the sport as it does about the randomness of a single game.

 

[dxmnkd316]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Even if you include regional games you’re still talking about a one and done scenario where even in the most lopsided matchups the underdog probably has a 30% chance of winning. The results of a one and done won’t tell you nearly as much as a sample that’s 40x greater.

I don’t think AIC winning a single game tells us as much about the parity of the sport as it does about the randomness of a single game.

The only thing that’s telling is the number of first time winners and variance of teams in the last ten years.

 

[Koho]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Sorry but the one and done flukey tournament results don't really prove anything other than the fact that a single game of hockey can have a ton of variance in outcomes. If you want to argue that the blue bloods are missing the tournament more often or backing your assertion with something that has a significant sample size then that'd be reasonable (and I haven't looked into it but that argument might actually have some legs to it).

If your larger point is that there's more parity in college hockey, then I agree. But that's probably the worst way to measure/test that hypothesis.

I was explaining that the parity has resulted from these changes would explain the trend in 1's losing to 4's, an often discussed statistic this time of year. I didn't claim to use it as a statistical analysis to support my hypothesis.

I would be interested in seeing if my memory matches the statistics. It seems like prior to around 2006, it was much less likely for a number 1 seed to lose to a 4. Or to look at 4 and 1 seeds overall winning percentage for the tournament before and after that period. Last year someone showed that number 4 seeds since 2006 really weren't far below .500. It used to seem like the first round was mostly a foregone conclusion in a number of matchups before that, and that you wouldn't see a 4 winning it all.

As for discarding stats based on a one and done tourney, lets keep in mind there are 15 games in the NCAA per year, so that is 180 games since 2006, not a bad sample size. (Every game played in the season is one and done too. Until the next game. They don't become more unpredictable at tourney time. Impact of one and done only matters if your sole analysis is who is the Champion.)

So the point is there seems to be a lot more parity following the CBA and resulting ramping up of the recruiting wars for younger kids, and I believe most people would agree that it is represented not only during the season, but also in the NCAA results for a number of years.

 

[Koho]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

I was explaining that the parity has resulted from these changes would explain the trend in 1's losing to 4's, an often discussed statistic this time of year. I didn't claim to use it as a statistical analysis to support my hypothesis.

I would be interested in seeing if my memory matches the statistics. It seems like prior to around 2006, it was much less likely for a number 1 seed to lose to a 4. Or to look at 4 and 1 seeds overall winning percentage for the tournament before and after that period. Last year someone showed that number 4 seeds since 2006 really weren't far below .500. It used to seem like the first round was mostly a foregone conclusion in a number of matchups before that, and that you wouldn't see a 4 winning it all.

As for discarding stats based on a one and done tourney, lets keep in mind there are 15 games in the NCAA per year, so that is 180 games since 2006, not a bad sample size. (Every game played in the season is one and done too. Until the next game. They don't become more unpredictable at tourney time. Impact of one and done only matters if your sole analysis is who is the Champion.)

So the point is there seems to be a lot more parity following the CBA and resulting ramping up of the recruiting wars for younger kids, and I believe most people would agree that it is represented not only during the season, but also in the NCAA results for a number of years.

Well I didn't read all the crap that came after the post I was responding to. Expected a statistics lesson from SteveO. And I am not going to get into that. Besides, it is the interweb, so I can draw a conclusion from 3 games if I want. But plenty of us who have watched the game for decades have seen the change occur just by looking at results of the way the NCAA brackets have been playing out. (I have run the company brackets where I work (thus my current focus is on NCAA games) and it sure seems like the trend is there. (Yes, you can always get better correlation with better sample sizes, but that doesn't mean trends seen with smaller sample sizes don't exist.) In contrast, not that I have done an ounce of stats on it, but the HS bracket seems to be more predictable in the first round than NCAA has become. But since some of you want to question the basis of my statements, I am open to seeing a statistical analysis that proves me incorrect. In the absence of your own data, why do you need to question one statement I made that, however small, does seem to support it?

 

[trixR4kids]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

180 games for hockey is significantly smaller than 35-40+ game seasons (Ivy schools tend to have less games) over 14 seasons for 58-60 teams, if you’re trying to figure out who’s truly the best teams and measure the parity of the sport. That’s all I’m saying. A team like AIC beating SCSU shows how random a single game of hockey can be but not a whole besides that. It happening a few times over 14 seasons isnt that shocking either.

And I’m not really arguing against the idea of parity in the sport, I’m sure it’s a lot better than it was in the 80’s if I had to guess. I just think some coaches adjusted to the 2005 draft rule changes better than others.

 

[The Rube]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

There is a lot more parity, CBA/etc contributing to it. In the past, the Blue Bloods got the star players, who would play 2-4 years, and then move on. The others would get the workhorses that had to try a little harder to make it. They gelled as a team.

Now? Those workhorses are beating the big dogs, because the big dogs have 4-5 star players, and then everyone else. These smaller teams have a full team full of "just good enoughs."

A common refrain about ASU in the last part of the season was that they were missing their best player. Well, teams like AIC, RIT, etc, don't really have a best player (as in super-standout). They ALL have to be pretty good, or at least good enough. Put them in a one-and-done? Yep, anything can happen.

 

[SteveO]

Re: Minnesota Golden Gopher Season 2018-2019: Fire Motzko

Well I didn't read all the crap that came after the post I was responding to. Expected a statistics lesson from SteveO. And I am not going to get into that. Besides, it is the interweb, so I can draw a conclusion from 3 games if I want. But plenty of us who have watched the game for decades have seen the change occur just by looking at results of the way the NCAA brackets have been playing out. (I have run the company brackets where I work (thus my current focus is on NCAA games) and it sure seems like the trend is there. (Yes, you can always get better correlation with better sample sizes, but that doesn't mean trends seen with smaller sample sizes don't exist.) In contrast, not that I have done an ounce of stats on it, but the HS bracket seems to be more predictable in the first round than NCAA has become. But since some of you want to question the basis of my statements, I am open to seeing a statistical analysis that proves me incorrect. In the absence of your own data, why do you need to question one statement I made that, however small, does seem to support it?

crap??? Hey now. BTW sorry about the stats lesson. I teach this stuff and my mentor in my PhD program earned a PhD in advanced stats from tOSU (back in the day of handheld calculators only ), so I'm hopelessly infected.

I agree (in bold) and that was my point about including regionals, even if it is one and done. I do accept student quantitative research projects with a minimum of 30+ participants. But I always stress at least 50-100 to reasonably decrease risk of a T-II error. At the end of the day, the two bottom lines in regression analysis are reducing the variance in prediction error and finding a reasonable trend line with salient limitations.

I'd be willing to crunch the data but with other projects I'm working on, I don't have the time to invest in it. Trix's argument of citing the inherent confound in the one and done format is also valid. IIRC last season's FF champ UMD had the worst W/L record of any Natty champ in history.

There is a lot more parity, CBA/etc contributing to it. In the past, the Blue Bloods got the star players, who would play 2-4 years, and then move on. The others would get the workhorses that had to try a little harder to make it. They gelled as a team.

Now? Those workhorses are beating the big dogs, because the big dogs have 4-5 star players, and then everyone else. These smaller teams have a full team full of "just good enoughs."

A common refrain about ASU in the last part of the season was that they were missing their best player. Well, teams like AIC, RIT, etc, don't really have a best player (as in super-standout). They ALL have to be pretty good, or at least good enough. Put them in a one-and-done? Yep, anything can happen.

And a lot times the undrafted "workhorses" try harder...all the time and build solid line chemistry. When I played minors there were always these guys that played a very intense physical game and/or sometimes cheap shot to get the scouts attention. We see a reflection of that effort from some of the smaller schools who are loaded with 4 year players now. Good thoughts Rube.

This recent NY Times article says some of the same things --> ARTICLE