CSNbbs
Season outlook - Printable Version

+- CSNbbs (https://csnbbs.com)
+-- Forum: Active Boards (/forum-769.html)
+--- Forum: CAAbbs (/forum-676.html)
+---- Forum: CAA Conference Talk (/forum-677.html)
+----- Forum: Drexel (/forum-681.html)
+----- Thread: Season outlook (/thread-862035.html)

Pages: 1 2


RE: Season outlook - J.B. - 11-06-2018 08:17 PM

(11-05-2018 11:47 PM)Water Boy Wrote:  This is how ESPN rates our chances to win in the OOC. A little more optimistic than HiroshimaCarp but not by much.
EMU - 13.2%
Rutgers - 14.3%
Bryn Athyn - 83 %
LaSalle - 18%
Boston U - 57.1%
Bowling Green - 27%
Loyola - 54.5%
NJIT - 45.7%
Robert Morris - 55%
UMBC - 47.2%
Quinnipiac - 43.4%
UConn - 12.5%
Temple - 15.9%

Only an 83% chance to beat Bryn Athyn? Holy Crap? And they percentages for the other games are saying that not much is expected of this team this year.


RE: Season outlook - dan10 - 11-07-2018 02:46 PM

And don't look now but our best D1 chance was against Boston U, who just beat Northeastern last night. Oof.


RE: Season outlook - DrachenFire - 11-08-2018 12:45 PM

I think we slip back to single digit wins this year, but hope (and this is solely based off of the Bickerstaff commitment) this year represents The Bottom.


RE: Season outlook - Dragon For Life - 11-08-2018 11:31 PM

Get on the Spiker Train! I'm predicting a big improvement by end of the year Drexel will be in the top 6.


RE: Season outlook - Dragon For Life - 11-13-2018 11:30 PM

As of now it looks like Coletrane and Matej Juric will be redshirted baring injuries to the rest of the roster.


RE: Season outlook - J.B. - 11-17-2018 07:24 PM

(11-05-2018 11:47 PM)Water Boy Wrote:  This is how ESPN rates our chances to win in the OOC. A little more optimistic than HiroshimaCarp but not by much.
EMU - 13.2%
Rutgers - 14.3%
Bryn Athyn - 83 %
LaSalle - 18%
Boston U - 57.1%
Bowling Green - 27%
Loyola - 54.5%
NJIT - 45.7%
Robert Morris - 55%
UMBC - 47.2%
Quinnipiac - 43.4%
UConn - 12.5%
Temple - 15.9%

ESPN might want to re-think these percentages as Drexel might turn out to be a better team than the on-paper pre-season outlook might suggest. Drexel was given just an 18% chance to win today, and they did. But the fact that they were only given an 83% chance to be a very bad Division III team and smacked them by 77, suggests that maybe these percentages might turn out to be just a bit wrong.


RE: Season outlook - dan10 - 11-17-2018 07:39 PM

In fairness their rankings are based off of last years poor season in which we lost our best players. Our team was very much an unknown and difficult to predict. Hell, we are 4 games into the season and most of us still dont know what to expect. So far we have exceeded expectations, but its too early to tell whats an anomaly and whats real. This should still be a team we should be able to get fans behind win or lose. This isnt Bruiser ball. Even when shots dont fall and they dont play particularly well on offense or defense, they play an exciting brand that should be appealing to college students.


RE: Season outlook - Water Boy - 11-17-2018 07:46 PM

The chance of us winning has climbed slightly before the game today, but it was still only 20 something. Probably the bigger surprise was how wrong Vegas has it with us being a 10.5 pt underdog.

I do think we are better that what the computers (Espn, KenPom, etc) think right now. You can't really blame them though as we have so many new pieces. 4 out of the 5 starters today didn't play for us last year. We might have some more dud games like Rutgers, but I'm excited for what's to come this year (and beyond).


RE: Season outlook - hiroshimacarp - 11-18-2018 04:14 AM

Spiker is a magician if he turns this into close to a .500 team. I was really starting to feel bad for him between the transfers and then all the squash news. Probably felt like he took a dead end job. I guess the contract extension helped.

It is too early to draw conclusions though. This is the same Lasalle outfit that has always been the punch line of jokes around here. Beating them should be an expectation especially when they’re rebuilding.


RE: Season outlook - J.B. - 11-18-2018 09:27 AM

Drexel is not at the point yet where beating ANY Big 5 team is an expectation. Except for about 2 years, La Salle has been rebuilding for about 23 years now. However, they are still in the Atlantic 10, which is a historically much stronger conference than the CAA. A-10 teams have more resources to get better players and succeed. Even though La Salle fits the footprint of a mid-major and has clearly been in over their heads in the A-10...they're still in the A-10. Therefore, any win for Drexel against them is a really big win.


RE: Season outlook - hiroshimacarp - 11-18-2018 04:22 PM

It’s a big win in the sense that I didn’t think we would beat anybody this year. That’s why I said “should” be an expectation. I have no expectations for this year. I come here to avoid spending money since the internet is just a big shopping mall now.

Temple is an interesting game in December. I’d like to see that we’re at least approaching the Bruiser days where we make noise in the city.


RE: Season outlook - Dragon For Life - 11-18-2018 09:13 PM

(11-18-2018 04:22 PM)hiroshimacarp Wrote:  It’s a big win in the sense that I didn’t think we would beat anybody this year. That’s why I said “should” be an expectation. I have no expectations for this year. I come here to avoid spending money since the internet is just a big shopping mall now.

Temple is an interesting game in December. I’d like to see that we’re at least approaching the Bruiser days where we make noise in the city.

Really.. not one win?


RE: Season outlook - Water Boy - 11-23-2018 12:30 PM

(11-17-2018 07:24 PM)J.B. Wrote:  
(11-05-2018 11:47 PM)Water Boy Wrote:  This is how ESPN rates our chances to win in the OOC. A little more optimistic than HiroshimaCarp but not by much.
EMU - 13.2%
Rutgers - 14.3%
Bryn Athyn - 83 %
LaSalle - 18%
Boston U - 57.1%
Bowling Green - 27%
Loyola - 54.5%
NJIT - 45.7%
Robert Morris - 55%
UMBC - 47.2%
Quinnipiac - 43.4%
UConn - 12.5%
Temple - 15.9%

ESPN might want to re-think these percentages as Drexel might turn out to be a better team than the on-paper pre-season outlook might suggest. Drexel was given just an 18% chance to win today, and they did. But the fact that they were only given an 83% chance to be a very bad Division III team and smacked them by 77, suggests that maybe these percentages might turn out to be just a bit wrong.
ESPN has made adjustments (I'm sure people at ESPN headquarters have lost a lot of sleep over it)
Bowling Green - 27% now 32%
Loyola - 54.5% now 63.4%
NJIT - 45.7% now 57.5%
Robert Morris - 55% now 61.7%
UMBC - 47.2% now 55.3%
Quinnipiac - 43.4% now 57.1%.
UConn - 12.5% now 14.5%
Temple - 15.9% now 22%

Strictly going by the percentages, this would suggest finishing out the OOC at 4-4 (28%) is the most likely outcome, followed by 3-5 (26%) and 5 -3 (18%). The likelihood of getting at least four wins which would give us a winning record going into CAA play is 53%.

Hopefully we are still being under-estimated a bit. Based on the original projections, we only had a 21% chance of having 3 wins at this point so one could see where that might be the case.


RE: Season outlook - EvanJ - 11-28-2018 10:07 AM

How did you do the probabilities of each record? If you treat the games as independent events, there are 2^8 = 256 possible outcomes. For each outcome, you would have to multiply 8 probabilities and then add up the outcomes that resulted in the same record. That's possible, but did you spend that long doing it in Excel? Given that the mean of the probabilities is 45.4%, it makes sense for 4-4 to be most likely followed by 3-5 followed by 5-3, but the mean of the probabilities doesn't tell you the probability of each record. Furthermore, probabilities do not have to be in the same order as their distance from the mean. For example, take 8 independent events with a probability of 0.2. The mean is 1.6, and 1.6 is closer to 3 than to 0, but 0 successes (p = 0.1678) are more likely than 3 (p = .1468). I'm not trying to be rude or question your knowledge of math, but I'm wondering how you did it. I could do probabilities for Hofstra, which would be quicker because Hofstra has only 6 Division I OOC games left, so there are only 64 possible outcomes.


RE: Season outlook - Water Boy - 11-28-2018 10:51 AM

(11-28-2018 10:07 AM)EvanJ Wrote:  How did you do the probabilities of each record? If you treat the games as independent events, there are 2^8 = 256 possible outcomes. For each outcome, you would have to multiply 8 probabilities and then add up the outcomes that resulted in the same record. That's possible, but did you spend that long doing it in Excel? Given that the mean of the probabilities is 45.4%, it makes sense for 4-4 to be most likely followed by 3-5 followed by 5-3, but the mean of the probabilities doesn't tell you the probability of each record. Furthermore, probabilities do not have to be in the same order as their distance from the mean. For example, take 8 independent events with a probability of 0.2. The mean is 1.6, and 1.6 is closer to 3 than to 0, but 0 successes (p = 0.1678) are more likely than 3 (p = .1468). I'm not trying to be rude or question your knowledge of math, but I'm wondering how you did it. I could do probabilities for Hofstra, which would be quicker because Hofstra has only 6 Division I OOC games left, so there are only 64 possible outcomes.

I did it in Excel but using a more empirical approach. I used the random number generator to assign a number between zero and one for each game. If that number was less than the probability of winning, I assigned a one, otherwise a zero. Then summing up over all the games, you get the number of wins for that realization. I then just copied down 10000 time to have that many realizations. Then I counted how many out of the 10000 realizations had the possible different win totals.

Hopefully that makes sense. It didn't take more than a few minutes to do. Probably the most time consuming thing was to click around in ESPN to get the win percentages. The numbers change slightly every time you calculate, but not by much. One could use more realizations if you wanted more accuracy.


RE: Season outlook - EvanJ - 11-28-2018 04:28 PM

I understand what you did, and it will be close enough. I did it in Excel in under 19 minutes, and that was with mistakes that slowed me down. Here are the probabilities:

8-0: 0.07%
7-1: 1.09%
6-2: 6.26%
5-3: 17.94%
4-4: 28.66%
3-5: 26.73%
2-6: 14.50%
1-7: 4.24%
0-8: 0.52%

It adds up to 100.01% due to rounding. Your simulations were correct to the nearest percent for 5-3. For 4-4 and 3-5, you were under 1% off, but rounding to the nearest percent makes 1% higher than your simulations.