PDA

View Full Version : Hockey East - Who's in, who's out, who's home: by the numbers - 2011-12 edition



Pages : 1 2 3 [4]

Todd
03-03-2012, 03:03 AM
Before getting to the numbers which, with one game left, are simple enough that most folks can do them on their own, I thought I'd start with an example of why clear tie-breakers matter.

Tonight, something happened at Agganis that you don't see very often. No, not 3-on-3 hockey. Well, actually, yes, 3-on-3 hockey, but that's not what I meant. What also happened was a coach pulling a goalie with over five minutes left in the game. Granted his team was down two goals, but ... over five minutes? How does that risk reward curve work?

It works like this.

Coming into the night, UMA and NU were tied for the final playoff spot. The tbs broke such that if BC or BU were the 1 seed, UMA won the tb; but if UML won the top seed, NU earned the tb.

With five minutes left, NU was down 4-2 against BU. UMA had gone up 3-1 at Mullins. BC was up 4-1 at Conte.

With a UMA win, that would put UMA two points ahead. A Northeastern loss would mean that the best the Huskies could do would be to draw back even with the Minutemen tomorrow.

If BC wins, they lock up a RS co-champ and can only be caught by BU (making their range 1-2). That BC win would mean that UMA would win a tb with NU, whichever team (BC or BU) ended up as the 1 seed.

If the other scores hold, that means that NU had to earn at least a point tonight to keep their season alive.

Those other scores did hold and NU was unsuccessful in closing the gap. NU loss. 9th place locked up. UVM now knows who their golfing buddy is.

-----
And now... a little recent history...

As it turns out, for this year's NU and BU senior classes, one of them has ended the other's season in their sophomore, junior and senior years.

This year, BU ended NU's playoff hopes at Agganis on BU's Senior Night with one game to play - actually the earliest either has nixed the other in the skein. The two squads face off tomorrow at Matthews for NU's Senior Night and then the Huskies can clear out their lockers. (In a scheduling quirk, tomorrow will be BU's third straight Senior Night game, including UVM at Gutterson last Saturday. By comparison, BC will play their first Senior Night game tomorrow. Their own.)

At the end of last season, NU and BU played five straight games, the latter four at Agganis. Like this year, they played a pair to end the RS. Following that, they played a best 2-of-3 at BU in the HE QFs. The road team won four of the five games, with the exception being BU taking the second QF game to extend the series another night. Since opening Agganis, BU has hosted the QFs every year, and every year it has gone to a third game. The 2010-11 Huskies were the first visitors to take Game Three. That 3-of-4 streak by the Huskies not only knocked the Terriers out of the HE tourney, it also knocked them off the NCAA bubble and, more suddenly than expected, into their off-season.

In 2009-10, NU finished up a roller-coaster second half with a pair against BU. After running off six straight league wins to put themselves in the thick of the playoff chase, NU ended the season 0-4-1 against BC, UNH and then BU. In perhaps the tightest race in league history, the difference between 3rd seed and 9th place was four points. Unfortunately for NU, they were that 9th place team. The sweep by BU put them atop the tie for 3rd. The losses by NU left them one point behind UVM for 8th. That was also the season that HE had their only three-way tie, plus another pair. Final tally: BU, ME, UML at 28, MC and UMA at 26, UVM at 25, NU at 24 and out. Considering that 27 points is a .500 record in HE, like someone trying to fit into jeans they've outgrown, that's some pretty tight bunching around the middle. That BU-NU series also impacted the NCAA tourney, even though neither team made it. By NU losing, UVM was #8 and played #1 UNH in the QFs. The Catamounts beat the Wildcats (I was tempted to say the Cats beat the 'Cats) in three games and went on to face #2 BC in the Semis. Helped by the strength of those four extra games against highly ranked teams - and their strong non-conf record - UVM made the NCAA tourney as HE sent their 1, 2 and 8 seeds to the Big Dance.

As freshmen, these Terriers and Huskies also had a couple of rarely-mentioned connections.

First of all, it is well documented that that BU team won almost everything they could that year. Every pre-, mid- and post-season tourney they entered for seven team trophies. They won the national Rookie, Player (Hobey) and Coach (Penrose) of the year awards. They even had a second player in the three Hobey Hat Trick finalists. What they didn't win was because of the third member of the Hobey Hat Trick, standout NU goalie Brad Thiessen. Despite having a much-heralded roster including the Hobey Baker winner and a runner-up, a BU player did not win Hockey East's Player of the Year. Thiessen did. In fact, NU also took league awards for Coach, Three Stars, Defensive D, Defensive F, and Goalie - which is six of the ten league awards. For as good as that Terrier team was, that Husky team was nearly their equal. They were separated by only a point in the final standings, and when they met for a H&H with only three weeks left in the RS, both games were ties. Even though their Beanpot, HE and NCAA tourney stumbles hide it somewhat, that NU team was probably the best Husky squad in recent memory - but they were overshadowed by the attention drawn by that season's Terriers. Where the Terriers managed to find a way to win, the Huskies always came up just short. History is unkind to silver medalists.

Secondly, many people know that that BU team beat every team on their schedule not from Hockey East (10-0-0). What I haven't heard from anyone else is the realization that they also ended the season of almost every other team in the league - and in order of the final standings, if not quite the seeding. On the final night of the 08-09 RS, BU beat the #9 PC team that missed the playoffs. In the QFs, BU knocked of #8 Maine while #2 NU finished #7 UMA. #6 BC lost to BU in the HE semis and #5 UML fell in the HE final. In the NCAA Regional, BU outlasted UNH who was tied for 3rd. In the FF Semis, they took out the other #3, UVM. Coming up the other side of the bracket, NU was leading Cornell before a closing burst by the Big Red overtook them - but that happened a lot in that all-around incredible tournament. Cornell then lost to Bemidji, who lost to Miami, who lost in the final to BU. Had NU continued, they were slated to meet BU in the final - where they would have each ended the other's season. One a champion and one a runner-up. Would BU have finished their progression through the HE standings, or would NU have gotten revenge for BU clipping them for the RS title on the last day of the season? We'll never know, but it would have been a heck of a game. It would have to wait for BU to knock off NU at the beginning of the next playoff race, as noted above. Instead, the BU/Miami final ended with a screened goalie and an unusual deflection and a BU national title.

Which brings us back to tonight.

The last time I recall seeing a coach pull the goalie with over three minutes to go was in that 08-09 national final. Similarly, BU was down by two with three-and-a-half to go and the season on the line. It worked out better for those Terriers than for tonight's Huskies, but I think both coaches were right to make such a gutsy call. Might as well go down swinging rather than leaving the bat on your shoulder. As Wayne Gretzky once said, "I miss 100% of the shots I don't take."

The other connection between that title game and tonight? The freshman screening the goalie. He doesn't play for BU anymore. He plays for Northeastern. In fact, by joining the Huskies' roster, he may have become the first NU player to wear a NCAA title ring. His first visit back to play at Agganis since leaving the Terriers? Tonight. Senior Night. What would have been his Senior Night.

So even though Northeastern was eliminated tonight, the season finale will still mean something. With the home team's fate already sealed, this game can exist in a bubble for both sets of Seniors. They can just play one final game to cap off what has been a remarkably mutually-intertwined run.

And then, as is their destiny, one of them will pack up their things and go home.

Todd
03-03-2012, 05:41 AM
BC 37 - 39 [1-2]
BU 35 - 37 [1-3]
UML 33 - 35 [2-3]
--- Home Lock - 32 (ME) ---
ME 30 - 32 [4-5]
MC 29 - 31 [4-5]
--- Home Eligible - 30 (ME) ---
UNH 24 - 26 [6-7]
PC 23 - 25 [6-8]
UMA 22 - 24 [7-8]
--- In - 22 (NU) ---
--- Out - 22 (UMA) ---
NU 20 - 22 [9]
UVM 7 - 19 [10]

Remaining LEAGUE schedules:
BC - UVM
BU - @NU
UML - PC
ME - UNH
MC - UMA
UNH - @ME
PC - @UML
UMA - @MC
NU - BU
UVM - @BC


One buwwet weft?*

Final night.

Let's see what we know and what we still don't know.

With their win, BC locks up a share of the RS title. The only way they aren't solo champs is with a loss and a BU win, in which case the Eagles drop to 2 and BU is 1.

UML has locked up not only Home Ice, but at least a 3 seed. The only way they move from 3 is with a win and a BU loss, in which case the Hawks rise to 2 and BU is 3.

If neither of the two above scenarios happen, BU is 2.

The 4 v 5 QF series will be ME/MC. Only the seeding and location is up for grabs. MC trails by a point, but owns the tb. If they gain at all on Maine, the series will be at Volpe. If Maine holds them off, it's at Alfond.

Note of caution to the Warriors: Last year, BU was the country's final undefeated team and rose to #1 in the polls because of it. They then slowly drifted down until they dropped off the NCAA bubble and out of the post-season. Merrimack was this season's last undefeated team and rose to #1 in the polls because of it. They are now tied for 15th in the PWR and would lose that tie, making them effectively 16th. Taking autobids into consideration, instead of looking at being the last team in to the 16-team tourney, Merrimack is looking at being the last team out.

That said, Maine doesn't have a lot of breathing room either, at 10th in the PWR.

Taking a longer detour into the PWR for a moment...

If NU loses to BU in the finale, the margins are razor thin of whether they end up above or below .5000. If they drop below .5000, they are no longer a TUC. Every HE team in contention for a NCAA bid has at least two wins against NU. If NU drops off, the entire league may lose spots in the PWR by their TUC record dropping with NU's sudden absence. Since this will be NU's last game either way, their RPI movements will thereafter be dependent upon their strength of schedule, since their winning percentage won't change.

Here we would have a situation similar to Holy Cross a few years ago when it actually made sense for them to throw the AHA final in order to make their opponent a TUC. Long story short, it could have raised them up into the second tier as an at-large and AHA could have brought two teams to the tourney. In this case, it could make sense for BU to tie NU, rather than beat them. Sure, adding a tie isn't as good as adding a win for your TUC record, but keeping the two wins you already have is important too. Basically, just for the TUC comparison, is it better to keep a 2-0-1 record than it is to have an irrelevant 3-0-0 record - especially when the first two wins are currently being factored in?

In a weird way, BU and BC could be helping each other out. If BU ties (or loses), BC definitley gets to keep their 3-0-0 TUC vs NU that is surely helping the Eagles stay so high in the PWR. If BU wins, NU drops down to just under - within a rounding error of significant digits - but since NU played BC an extra time in the Beanpot, it actually helps them stay higher if BC beats UVM. At that point, the margin is close enough that how deep NU's non-conf opponents go into their post season could make the difference between TUC/No-TUC for the Huskies - and pair/no-pair for the rest of the top HE contenders. Worth keeping an eye on.

Conversely, UMA's win on Friday keeps them above the TUC bubble through the RS. Of course, if they get swept in the playoffs, they could drop off. That could leave either QF opponent BC or BU, in the unenviable position of facing a team that you knock off the TUC bubble if you beat them, but they stay if they beat you. In sum, your wins wouldn't count, but your losses would.

Further, for BU, the 1-1-1 record v UMA drags down their over-.5000 TUC record a little. For BC, who went 1-2-0 vs UMA, it drags it down a lot. For either team, it would make sense to win not only because you always want to win, especially in the playoffs, but also because in knocking off UMA, you wouldn't care so much about the new wins that you weren't adding, but the old losses/ties that you get to throw away.

OK, back to the HE standings...

1,2,3...
4, 5...
Ah, here we are...

UNH/PC/UMA.

UNH wins either or both tbs pretty handily, so they can't be caught by UMA. In fact, the outcomes of this group are similar to the top three in that UNH could be 6 or 7, UMA could be 7 or 8 and current meat-in-the-sandwich PC could be 6, 7, or 8. The difference is that the top trio is separated by pairs of points, but the bottom team has the tb, so there's only one point of wiggle room for the current higher seed. In this group, the division is only one point, but the higher team has the tb (UNH over PC, UMA and PC/UMA, PC over UMA - and UMA over NU), expanding their cushion to a single point (from none at all).

UNH's M# for 6th is 1 w/ PC. PC's M# for 7 is 1 w/UMA. In fact, it could be that we only need one result to seed this whole group. If PC ties (earning one point to clinch 7, but missing out on the second point from a win, clinching 6 for UNH), it's UNH 6, PC 7, and UMA 8.

NU, as noted in the pull-the-goalie section of the prior post, is now 9th.

UVM... still dead (http://www.hulu.com/watch/1426/saturday-night-live-the-death-of-franco) - but now they have company.


*
Elmer Fudd: Well, what do you know? One buwwet left (http://www.imdb.com/character/ch0029092/quotes)!
Bugs: One buwwet left? Hey, Laughing Boy, there was...
Daffy: [as his scalp, with a bullet lodged in it, flaps behind him] I know, I *know*!

Go-UML
03-04-2012, 09:28 AM
Todd, thanks again for the insight this season. Enjoy the postseason!

Todd
03-04-2012, 09:04 PM
BC 39 [1]
UML 35 [2]
BU 35 [3]
ME 32 [4]
--- Home Ice ---
MC 31 [5]
UNH 24 [6]
PC 23 [7]
UMA 22 [8]
--- Out ---
NU 22 [9]
UVM 7 [10]


QFs

UMA @ BC
PC @ UML
UNH @ BU
MC @ ME


At the moment, HE has five teams that have a shot at making the NCAAs, but it may be a matter of how many other league teams remain Teams Under Consideration (by having an RPI of .5000 or higher). With TUC being the most volatile of the criteria used for the PWR, late swings here often move teams drastically up and down as various results come in throughout the playoffs. The reason for the volatility is that as teams qualify - or become disqualified - as a TUC, their results against the contenders very suddenly matter or don't, and could help decide who does and does not make the tourney, or where they are placed.

An example of this can be seen in BU's 2007-08 (IIRC) NCAA fate. During the league playoffs, they were seeded as high as 10th in the PWR. However, most of their league TUC wins were against teams that faded down the stretch, specifically PC, NU and UML. At the time, the TUC criteria was not ">= .5000 RPI", but "Top 25 RPI". That meant that not only did each team's own RPI matter, but so did those RPIs around them. Long story short, due to things like the Quinnipiac/Harvard ECAC QF going to a third game and Air Force winning their league title, those three HE teams ended up 26, 27 and 29 in RPI, making all of them - and BU's 8-1-0 record against them - drop off the TUC list and taking BU's NCAA chances with them. (That's all from memory, so if someone finds an error, let me know.) Now, like every other team, it's no one's fault but BU's that they didn't make that tourney. The point here is to demonstrate how a team could drop from a comfortable 2 seed in the NCAAs to falling off the bubble, based primarily on whether or not various teams are or are not TUCs.

Currently, here are the TUC records of the five HE teams looking to the NCAAs
BC: 17-10-1 . 6250
UML: 13-7-1 .6429
BU: 16-10-1 .6111
ME: 11-10-3 .5208
MC: 9-8-6 .5217

Since all of these records are between .6667 and .5000, adding an opponent against whom you went 2-1-0 (.6667) or better brings your record up, but adding in a 1-1-1 (.5000) or worse drags your record down. For results already being included (against teams that are already a TUC), keeping quality outcomes keeps your record high, but a team dropping out means those results go away - and then your record drops. Similarly, losing a team against whom you have a less favorable record actually improves your chances.

With each team's own record, their oppponent's records and their opponent's opponent's record factoring in to RPI, there is a bit of small variation in RPI that I won't calculate because it would have to factor in every conceivable outcome for the rest of the playoffs for every team across the country. That said, the strongest influence on RPI is a team's own winning percentage. Simply calculating those changes allows for a machete-level estimation without needing the scalpel of factoring in the rest of the results around D1. The projections below use that broader level of certainty, rather than exacting precision.

Note that, ironically, most of the Hockey East results besides their own won't mathematically matter much for most teams. Since league opponents generally play each other three times each, one team winning and another losing essentially washes out in Strength of Schedule (SOS). The exceptions to that are when one HE team plays another in a non-conf tourney. In such cases, the opponent that they faced for an extra game carries more weight on their SOS. That influence does extend to the strength of the opponents that make it through the tourney, as mentioned in a prior post about 2009-10 8-seed UVM facing top seed UNH three times and then 2-seed BC. Those were four very stong opponents added to their SOS, which gave their RPI a bit of a kick.

So...

After being swept this weekend, PC has dropped to .4883 and would need to add three wins get back over .5000. At that point, they would be in the HE final. A loss there would drop them back under and out, so either PC wins the HE tourney or they won't be a TUC. If PC should move up, BC would add a 3-0-0 record. BU, UML and ME would each add 2-1-0. MC would add a 1-2-0 record. The Warriors, who are most vulnerable of the five at the moment, might not be able to handle adding in that record, so just as well for them if PC stays out.

With their win last night, NU jumped up to .5074. Since they're done playing, they won't add any more wins or losses of their own (there are no longer any ties in playoff hockey), they will only move with the slight variations in other teams' results. NU will stay a TUC and their current impact will stay as it is. Had they lost, they would have been right around the .5000 mark and that would have been among the more interesting numbers to watch as their TUC status would likely have depended on those minor variations - so, for example, whether Notre Dame (who faced NU twice) won or lost might affect the NCAA seeding for all of HE.

For UMA, starting at .5020, it gets interesting like it might have for NU. If they get swept by BC, they would be down around the .4950-.4960 area, plus or minus. However, if they take one of the three, they jump up to around .4995, just based on their own win %. The fact that those would be three extra games against an opponent with as strong an RPI as BC's would probably be enough to carry them over .5000 and keep them a TUC. Results around the country would impact their results as well, most notably: Bentley and Holy Cross from AHA and Quinnipiac, Harvard, Yale and Cornell from ECAC. Since UMA played Maine an extra time down in Florida, ME's playoff success, or lack thereof, comes into play here, too.

If UMA were to win two playoff games (or more) and then be eliminated short of the title, whether winning in the QF 2-1 or 2-0 and then losing at the Garden, they would definitely be a TUC.

If BC were to sweep UMA, here's what each team would lose: BC 3-2-0 (1-2-0 + 2-0-0 in QFs), UML 3-0-0, BU and ME 1-1-1, and MC 2-1-0.

BU and ME would clearly benefit by UMA going away, as dropping three games of .5000 would raise their TUC records. For BU, the change from 16-10-1 to 15-9-0 goes from .6111 to .6250. For Maine, 11-10-3 becomes 10-9-2, so .5208 would rise to .5238. (Since they're only a game over .5000, the effect of dropping .5000 is less drastic than for BU who is further from .5000.)

Looking more closely at BC's case, what is currently being counted is the 1-2-0. Sure, in sweeping UMA and knocking them off TUC status, the Eagles wouldn't get credit for the two new wins - which would put them at 19-10-1 or .6500 - but dropping the poor 1-2-0 from what they already have brings them to 16-8-1, or .6600. (For a quick estimate, since 3-2-0 is .6000, dropping it from .6500 - which is higher - has to improve things.)

In sum, based on 8-seed UMA's surprisingly good record against some of the top teams, it's actually helpful for BC, BU and ME for UMA to go away. For UML, it would drop them to .5833 and MC would drop to .5000. A .5000 TUC record is very dangerous to the team that holds it.

UNH is at .5010. If they were to get swept, they'd drop out. This would be huge for BC and BU, who are both 3-0-0 against UNH and that would go away. For UML and ME, they'd lose a 2-1-0. MC would actually benefit from their current situation, dropping a 1-1-1 - unless UMA had already dropped out, in which case, they'd already be at .5000, so losing a .5000 would literally have no impact.

Like UMA, if UNH were to win two in the first round, whether they end up with zero, one or two losses to go with that, they would stay above .5000 and stay a TUC.

Now, the 1-2 QF result for UNH is where it gets interesting. Based just on their own win %, UNH would drop to around .4985. However, factoring in the three more games against BU, they could be about .5004. Because of the extra three games between UNH and BU (six total), BU's fate would have a larger influence on then-idle UNH's SOS than other teams. Isolating that effect from all other results, if BU were to then lose in the semis, UNH's RPI would drop to around .5001. BU losing the final would make it almost .5003. BU winning HE would bring UNH to nearly .5006.

Clearly, this is the situation that bears the most watching. Also strongly influencing this razor thin margin would be Brown, Dartmouth, Harvard and Union from the ECAC. St. Cloud would have double weight, from the two games played against UNH. Even though Alabama-Huntsville doesn't have a post-season, the changes to their RPI from their schedule - which, by definition, is all non-conference since they're Independent - is also a factor for UNH.

Looking at both UMA and UNH's list of ECAC schools, some of those will cancel each other out if they face off in the semis and/or finals. Between UNH and UMA, one of them will drop slighty and the other will rise, because Brown (UNH's list) and Quinnipiac (UMA's list) are playing an elimination game right now.

Terrier520
03-04-2012, 09:28 PM
For BC, it's easy: Just keep winning. If they get to the semis (or championship), they're virtually guaranteed a 1 seed.

But I'm not sure if it's better for BU to go 4-2-0 against UNH (and lose the QF matchup), or go 5-1-0 and see UNH drop below TUC. But I think for BU the answer is just keep winning.

ETA: Thanks, Todd. Excellent post, as usual. I imagine you hibernate during the summer, since you don't seem to sleep now.

Priceless
03-04-2012, 09:46 PM
If BC were to sweep UMA, here's what each team would lose: BC 3-2-0 (1-2-0 + 2-0-0 in QFs), UML 3-0-0, BU and ME 1-1-1, and MC 2-1-0.

BU and ME would clearly benefit by UMA going away, as dropping three games of .5000 would raise their TUC records. For BU, the change from 16-10-1 to 15-9-0 goes from .6111 to .6250. For Maine, 11-10-3 becomes 10-9-2, so .5208 would rise to .5238. (Since they're only a game over .5000, the effect of dropping .5000 is less drastic than for BU who is further from .5000.)

Looking more closely at BC's case, what is currently being counted is the 1-2-0. Sure, in sweeping UMA and knocking them off TUC status, the Eagles wouldn't get credit for the two new wins - which would put them at 19-10-1 or .6500 - but dropping the poor 1-2-0 from what they already have brings them to 16-8-1, or .6600. (For a quick estimate, since 3-2-0 is .6000, dropping it from .6500 - which is higher - has to improve things.)

In sum, based on 8-seed UMA's surprisingly good record against some of the top teams, it's actually helpful for BC, BU and ME for UMA to go away. For UML, it would drop them to .5833 and MC would drop to .5000. A .5000 TUC record is very dangerous to the team that holds it.

Maine beat UMass in Florida, so is 2-1-1 vs the Minutemen. Losing them would give the Black Bears a 9-9-2 TUC record.

Patman
03-04-2012, 10:11 PM
additional nitpick... if lowell loses to PC then they certainly won't have a 2-1-0 record against them

Todd
03-05-2012, 02:10 AM
Maine beat UMass in Florida, so is 2-1-1 vs the Minutemen. Losing them would give the Black Bears a 9-9-2 TUC record.

Good catch.

I looked at the HE records, but should have re-calced ME/UMA, esp since I noted it in the other part of the UMA section.

If NU were in question, the BC/NU Beanpot game would be a similar note. Bc/BU in the BP is already accounted for and BU/NU didn't happen at the BP.

Todd
03-05-2012, 02:14 AM
additional nitpick... if lowell loses to PC then they certainly won't have a 2-1-0 record against them

Right. Best case they'd be 3-3-0 (adding 1-2-0 to 2-1-0), but might be 2-3-0.

UML better hope that PC doesn't sweep through to the league title...

Go-UML
03-05-2012, 02:30 AM
Right. Best case they'd be 3-3-0 (adding 1-2-0 to 2-1-0), but might be 2-3-0.

UML better hope that PC doesn't sweep through to the league title...Well, we definitely have a say in if that happens or not. :p