Announcement

Collapse
No announcement yet.

2016-2017 WCHA Season Thread

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Re: 2016-2017 WCHA Season Thread

    Originally posted by John J. MacInnes View Post
    They are off to a great start. And Ferris State made a pretty good argument in 2013-14 that you can get off to a great start and 'run & hide' from the chasing pack. But if Tech matches Bemidji's points this weekend and then sweeps the following week, the gap is 6 points - or 2 wins. It only looks like a huge lead because of the new point system.
    I agree to some extent. The real question is can Bemidji keep up the scoring defense?
    Michigan Tech Legend, Founder of Mitch's Misfits, Co-Founder of Tech Hockey Guide, and Creator/Host of the Chasing MacNaughton Podcast covering MTU Hockey and the WCHA.

    Sports Allegiance: NFL: GB MLB: MIL NHL: MIN CB: UW CF: UW CH: MTU FIFA: USA MLS: MIN EPL: Everton

    Comment


    • Re: 2016-2017 WCHA Season Thread

      Originally posted by gfmorris View Post
      Well, it's obvious that Tech has their future in their hands with the four games. If I carry the math to the end, I'm willing to bet that Tech, on average, finishes in the top two and sometimes finishes in the top spot. That's entirely possible. I just don't know how likely it is.
      Originally posted by gfmorris View Post
      My argument stands. If you presume that the Beavers will win about 3-in-4 (and I do) whether they be against Tech or not *it gets very hard for Tech to catch up. Even if Tech beats Bemidji four times, it's still a lot of ground to catch up.
      Dicey presumption for sure. How are you at investing based on your presumptions and math?
      Last edited by camilo; 11-14-2016, 11:48 AM.

      Comment


      • Re: 2016-2017 WCHA Season Thread

        Not that things are in full swing, the Tech Hockey Guide WCHA Web Media Poll is out:
        Check out the results of the first poll
        Wild guess who's number 1 :P
        Last edited by Shirtless Guy; 11-14-2016, 03:00 PM.
        Michigan Tech Legend, Founder of Mitch's Misfits, Co-Founder of Tech Hockey Guide, and Creator/Host of the Chasing MacNaughton Podcast covering MTU Hockey and the WCHA.

        Sports Allegiance: NFL: GB MLB: MIL NHL: MIN CB: UW CF: UW CH: MTU FIFA: USA MLS: MIN EPL: Everton

        Comment


        • Re: 2016-2017 WCHA Season Thread

          Originally posted by camilo View Post
          Dicey presumption for sure. How are you at investing based on your presumptions and math?
          If it's dicey, refute it.

          I let professionals manage the stock market. Also, these are probabilities. I wouldn't put BSU's chance of being the #1 seed much higher than 60% right now, but I would put them having home ice at something like 95%.

          Hey, I'm going to be fine if you all want to come around in mid-February and call me a dunce or something. That's fine.

          GFM
          Geof F. Morris
          UAH BSE MAE 2002
          UAHHockey.com

          Comment


          • Re: 2016-2017 WCHA Season Thread

            Originally posted by gfmorris View Post
            If it's dicey, refute it.

            I let professionals manage the stock market. Also, these are probabilities. I wouldn't put BSU's chance of being the #1 seed much higher than 60% right now, but I would put them having home ice at something like 95%.

            Hey, I'm going to be fine if you all want to come around in mid-February and call me a dunce or something. That's fine.

            GFM
            You probably know that most professional management of stocks can pretty much be explained by randomness.

            Now that I'm reading you more carefully, I don't care to argue at all because it all seems reasonable to me. Thanks. But the true test is long run beating simplistic predictions or WAGs - sort of like how investment managers should be measured, but that's a losing game for them.

            In doing this sort of prediction, what sort of factors have you used. Just some thoughts that come to my mind are: Have you considered whether the future schedule equal in difficulty to the past? Home vs. away balance, and which teams, where (good ones at home, bad ones away or vice versa)? Are potential injuries or other random challenges considered - viruses that can knock a team down for a weekend or two, potential academic issues, etc. How about the relative youth or experience of the team vs. teams they've played so far vs. in the future , and how does that factor statistically? Do experienced teams tend to improve less or more over the season than inexperienced teams?

            I have no idea what impact these factors would have. I don't know if you've accounted for factors you can fairly objectively predict or measure (strength of schedule, home vs. away, etc. past vs. future), or have plugged in some sort of probability of random events, so I ask. I don't claim that any of them refute your presumption, but wondering if they or others are factored in.

            In the a little knowledge can be dangerous department: I've taken some statistics and simulation courses in college (business school - operations management) about 25 years ago and it was fascinating, and very complicated and difficult for me, even at the extremely simple level I was able to "master". I had pretty good luck with my capstone project, because it involved a process I knew pretty well, and I was able to identify critical factors and gin up probabilities for enough of them to achieve good predictions and model for adjusting the process. I say "gin up" but really I tried to do sampling and probabilities/distributions within the limits of how much time and access I had at the time. A little better than a WAG.

            But I can't argue or refute with any degree of expertise in the field, which I assume you have. At this point, I'm a general manager (called CEO) of a small company working shallowly but broadly in the non-profit, for-profit and governmental worlds simultaneously. My primary expertise is tolerating uncertainty and being comfortable making imperfect decisions. That being the culture I work in, I always am skeptical of people who think they can calculate stuff like this but interested in how they do it.

            All that said, it's hard to disagree that BSU is the team to beat, but I don't claim to figure that out with Math, but just because they're a darn good team this year and only have a couple of teams in the league that can be said to be "good". It will be very interesting to see how they do in their 4 games with MTU.
            Last edited by camilo; 11-15-2016, 12:16 AM.

            Comment


            • Re: 2016-2017 WCHA Season Thread

              Originally posted by camilo View Post
              Whether you're right or wrong at any given point in any given season won't pursuade me that your model or whatever it is is actually correct. Let's see your modeling beats other guesses year after year to a significant degree. One season success in predicting a league championship can happen if enough people - or even one person - throws darts.

              As far as "refute", I just think it's too simplistic, unless I've missed what objective criteria you've used to make the presumption you stated, what factors, and on what weighting or probabilities you're including them. If so, I'm seriously interested, not being argumentative for its own sake.

              What I hear you saying is simply that past performance predicts future performance. Question's I'd ask (maybe you have) about your presumption are: Have you considered whether the future schedule equal in difficulty to the past? Home vs. away balance, and which teams, where (good ones at home, bad ones away or vice versa)? Are potential injuries or other random challenges considered - viruses that can knock a team down for a weekend or two, potential academic issues, etc. How about the relative youth or experience of the team vs. teams they've played so far vs. in the future , and how does that factor statistically? Do experienced teams tend to improve less or more over the season than inexperienced teams?

              I have no idea what impact these factors would have. I don't know if you've accounted for factors you can fairly objectively predict or measure (strength of schedule, home vs. away, etc. past vs. future), or have plugged in some sort of probability of random events, so I ask. I don't claim that any of them refute your presumption, but think they should be factored in. Maybe your simulation or model or whatever you call it accounts for these or more factors in a rigorous way. If so the dicey-ness factor I stated earlier is reduced! If so, again I'm truly interested and not just trying to be argumentative.

              In the a little knowledge can be dangerous department: I've taken some statistics and simulation courses in college (business school - operations management) about 25 years ago and it was fascinating, and very complicated and difficult for me, even at the extremely simple level I was able to "master". I can't argue or refute with any degree of expertise in the field, which I assume you have.

              All that said, it's hard to disagree that BSU is the team to beat, but I don't claim to figure that out with Math, but just because they're a darn good team this year and only have a couple of teams in the league that can be said to be "good". It will be very interesting to see how they do in their 4 games with MTU.
              He's basing this solely on his ELO rankings (adjusted for WCHA games only). ELO is one of the gold standards for odds and is used all over the place, FiveThirtyEight uses it for predicting odds on pro sports and it was created out of predicting winners in international chess.
              Michigan Tech Legend, Founder of Mitch's Misfits, Co-Founder of Tech Hockey Guide, and Creator/Host of the Chasing MacNaughton Podcast covering MTU Hockey and the WCHA.

              Sports Allegiance: NFL: GB MLB: MIL NHL: MIN CB: UW CF: UW CH: MTU FIFA: USA MLS: MIN EPL: Everton

              Comment


              • Re: 2016-2017 WCHA Season Thread

                To answer camilo's basic question: have I applied an Elo-based prediction model week-by-week in a Monte Carlo simulation to the end of the year? No, I haven't. I wish that I'd had the time to pull it together by now, but that hasn't happened. (My work hours suck, but I'll have some free time while watching the Earth go by over the weekend.) It would indeed be valuable.

                To put it cleanly, Bemidji has four potential sweeps: UAH, UAA, NMU, FSU. That gives them eight wins. As I noted above, if the Beavers go .500 down the stretch, it's still a situation where Tech has to win 13 in 20 to catch up. Is that possible? Yeah! But four of those games are indeed against Bemidji, and at this point, BELOW/ABOVE would argue that they teams would split those games — and Tech sweeping Bemidji would be something like flipping heads four times in a row (although the probabilities favor Tech with every win).

                It's important to note here that BELOW is simply an estimate of teams' strength. Tech is really high because they were really high last year and have played decently well to-date. Bemidji is high because BOOM! Seriously, they're +160 on the year. BELOW is saying, "Hey, we thought this was a .500 team, but really it's a .740 team." Is Bemidji that good? Well, they're playing well above that right now. I'd say that the truth is somewhere between .500 and .950, so .740 seems as good of a fit as any.

                That said, Monte Carlo simulations with probabilistic estimates are only as good as their inputs and assumptions. We have a pretty small data set — 140 games for 10 teams — so I'd probably estimate that the error is anywhere between 0-50 points, and again, we're dealing with probabilities —*a 75% chance that Bemidji wins any average game is still a 25% chance that they'll lose. That's about the odds that FiveThirtyEight had on the last election, and the lesser odds proved true. [Note: no desire to discuss politics.]

                GFM
                Geof F. Morris
                UAH BSE MAE 2002
                UAHHockey.com

                Comment


                • Re: 2016-2017 WCHA Season Thread

                  I can say with 98% confidence that Geof's blood pressure rises a little bit every time he has to type:

                  B-E-M-I-D-J-I
                  or B-E-A-V-E-R-S
                  or B-S-U

                  Millsy

                  NCAA TOURNAMENT 2005, 2006, 2009, 2010!
                  FROZEN FOUR 2009!


                  "Like" The BeaverPond's Facebook Page

                  BEMIDJI STATE BEAVERS!
                  NAIA National Champs: 1968, '69, '70, '71, '73, '79, '80
                  NCAA D-III National Champs: 1986
                  NCAA D-II National Champs: 1984, '93, '94, '95, '97
                  NCAA D-I National Champs: 20??
                  Perfect 31-0 in 1983-84
                  Holders of an NCAA Mens Record 43 straight wins (Nov. 8, 1983-Jan. 1, 1985)

                  Comment


                  • Re: 2016-2017 WCHA Season Thread

                    So, BSU's pretty good this year? Fun. Hi to all the old timers. Gonna try to catch one of the BSU @ BGSU games, if anyone will be around in January. Sorry to bother everyone - carry on...
                    Current NCAA D-I rinks I've been to:

                    AHA:
                    B1G: UMinn, UWisc
                    ECAC:
                    HEA: UMass, Notre Dame
                    NCHC: UMD, UND, SCSU, WMU
                    WCHA: UAH, BSU, FSU, LSSU, MSU, MTU, NMU

                    Inactive: BSU, UMD, UND, NMU, Notre Dame

                    Comment


                    • Re: 2016-2017 WCHA Season Thread

                      Originally posted by Shirtless Guy View Post
                      Not that things are in full swing, the Tech Hockey Guide WCHA Web Media Poll is out:
                      Check out the results of the first poll
                      Wild guess who's number 1 :P
                      Classic reason showing why pre-season polls mean squat.

                      Comment


                      • Re: 2016-2017 WCHA Season Thread

                        Originally posted by moose97 View Post
                        So, BSU's pretty good this year? Fun. Hi to all the old timers. Gonna try to catch one of the BSU @ BGSU games, if anyone will be around in January. Sorry to bother everyone - carry on...
                        Hey Moose, good to see ya back! Hopefully you can get to one of those games.

                        Man, do I miss The Hockey Show with you and Scotty........ Always got me through Fridays at work.
                        Millsy

                        NCAA TOURNAMENT 2005, 2006, 2009, 2010!
                        FROZEN FOUR 2009!


                        "Like" The BeaverPond's Facebook Page

                        BEMIDJI STATE BEAVERS!
                        NAIA National Champs: 1968, '69, '70, '71, '73, '79, '80
                        NCAA D-III National Champs: 1986
                        NCAA D-II National Champs: 1984, '93, '94, '95, '97
                        NCAA D-I National Champs: 20??
                        Perfect 31-0 in 1983-84
                        Holders of an NCAA Mens Record 43 straight wins (Nov. 8, 1983-Jan. 1, 1985)

                        Comment


                        • Re: 2016-2017 WCHA Season Thread

                          Originally posted by beaverhockeyfan View Post
                          I can say with 98% confidence that Geof's blood pressure rises a little bit every time he has to type:

                          B-E-M-I-D-J-I
                          or B-E-A-V-E-R-S
                          or B-S-U

                          You know me well, friend. But it's been 15 years, sooooooooooo ...

                          GFM
                          Geof F. Morris
                          UAH BSE MAE 2002
                          UAHHockey.com

                          Comment


                          • Re: 2016-2017 WCHA Season Thread

                            The short version of this discussion is that Tech and BSU play in three weeks. If Tech wins four games in AK and then sweeps the Beavers, Tech is first place based on a tie breaker (head to head). Not that it really matters in November, but Tech controls their own destiny to be back in first place after the first weekend in December.
                            Ryan J
                            Preserving Michigan Tech's Hockey History
                            https://www.johnsonsjerseys.net
                            Originally posted by geezer
                            Tech has the best of everything, even the best jersey nerd.
                            Originally posted by manurespreader
                            ...I really enjoyed listening to Ryan Johnson. He sounded intelligent.

                            Comment


                            • Re: 2016-2017 WCHA Season Thread

                              Originally posted by JohnsonsJerseys View Post
                              The short version of this discussion is that Tech and BSU play in three weeks. If Tech wins four games in AK and then sweeps the Beavers, Tech is first place based on a tie breaker (head to head). Not that it really matters in November, but Tech controls their own destiny to be back in first place after the first weekend in December.
                              Ryan J
                              It's true but the Alaska trip is always tough and BSU has played us tough for as long as I've been following Tech, so I would be very surprised to see that result.
                              2005-2006: 7-25-6
                              2006-2007: 18-17-5
                              2007-2008: 14-20-5
                              2008-2009: 6-25-7
                              2009-2010: 5-30-1

                              Comment


                              • Re: 2016-2017 WCHA Season Thread

                                Originally posted by JohnsonsJerseys View Post
                                The short version of this discussion is that Tech and BSU play in three weeks. If Tech wins four games in AK and then sweeps the Beavers, Tech is first place based on a tie breaker (head to head). Not that it really matters in November, but Tech controls their own destiny to be back in first place after the first weekend in December.
                                Ryan J
                                yes and if a submarine had wings it would fly too, but I'm not holding my breath.
                                MTU: Three time NCAA champions.

                                It never get's easier, you just go faster. -Greg Lemond

                                Comment

                                Working...
                                X