Like I said, I have just as many "facts" that paint Bailiff in a positive light, but for the sake of my mental health, please excuse me from this discussion.

(12-29-2015 09:14 AM)Afflicted Wrote:  Like I said, I have just as many "facts" that paint Bailiff in a positive light, but for the sake of my mental health, please excuse me from this discussion.

What facts? And simply saying he has taken us to 3 bowl games and won a conference championship doesn't cut it, as we've already shown that even in those years our "success" was largely based on beating teams ranked outside the Top 100.

(12-28-2015 08:50 PM)Afflicted Wrote:  

(12-28-2015 08:35 PM)Tomball Owl Wrote:  What's messed up about Antarius' data? As you said the data were facts.

Antarius dwells on the negative. His calling is to destroy Bailiff. I could list several "facts" that make Bailiff look great. It doesn't necessarily mean that he is "great," it just means that he has done some "great" things. So, yes, under Bailiff Rice has beaten some poor competition. At the same time, his teams have mostly taken care of business against those bad teams and advanced to three bowl games, and won two of them. You see what I mean?

Getting to and winning a bowl game is no big deal. Half of the competition gets to bowl games. Losing teams get to bowl games. Bailiffs only meaningful accomplishment in nine years is beating Marshall for the conference championship.

David Bailiff is a better head football coach than Bo Hagan was.

Spam is better than dog food.

quote='Owl 69/70/75' pid='12837736' dateline='1451401434']
David Bailiff is a better head football coach than Bo Hagan was.
[/quote]

It's been so long since we were in a conference that I cared about.
I, for one, look forward to any analysis, like we're seeing in this thread, of our play back in the SWC era.
Yes, 0-11 can't be made to not suck.
Why not take the analysis to the beginning of Rice football?
If we're going to flog a dead horse, flog it thoroughly.

(12-29-2015 11:23 AM)Grungy Wrote:  It's been so long since we were in a conference that I cared about.
I, for one, look forward to any analysis, like we're seeing in this thread, of our play back in the SWC era.
Yes, 0-11 can't be made to not suck.
Why not take the analysis to the beginning of Rice football?
If we're going to flog a dead horse, flog it thoroughly.

Sagarin online only dates back to 1995 I believe, the Hatfield era.

I guess, IMO, how we used to be or how Bailiff stacks up against Bo Hagan isn't really important. I think we all agree that Rice has been a rudderless tanker listing in the currents for decades. So, even if we do prove that Bailiff is better than his last 5 predecessors, it just shows that we have had 6 lousy coaches, as its clear from this data and otherwise that Bailiff cannot take Rice any further up the chain.

If I get time, I might plot Rice vs. "peer" coaches. For example, other G5 coaches within a standard deviation of salary, for example

Next, for more granularity, here is wins and losses vs Sagarin for all games



The x axis is game number with 1 being the first game of 2009. When I get a chance (post landing etc) I will try to clean it up. Based on this graph, we see that our losses are nearly flat while the teams we beat are getting increasingly worse.

Another interesting plot. Sagarin vs our score differentials in all games

What stands out here is a clear cluster that further delineates the “Bailiff line”, which is around the top 100 teams. Additionally our worst two losses came in 2015

Ok now your plots are starting to look good. I like these two (48&49).

Easy suggestion #1:
Plot losses (margin of loss) shown in post #49 as negative values below axis. That will show the (few) wins against better teams above all the losses and give a better view of the crossover point (we usually beat teams above x, we rarely beat teams above y)

Hard suggestion #2:
On the post 48 graph, plot margin of victory as an extended vertical bar above (for a win) or below (for a loss). That gives some sense of how we trend historically playing against different levels.

Suggestion 3: if you have to use Sagarin, use his predictor raw score. It roughly equates to projected point differential, so the vertical scale is consistent for the extended vertical bars in suggestion 2. The result would be a trend line that might compare to the performance ratings produced by Sadow, but with the additional useful context of how good the opponent was and what the scoreboard result was each game.

I would suggest that in #48, the y-axis be inverted, going up from the higher Sagarin numbers (worse teams) to the lower numbers (better teams).

That would show most of our victories on the bottom of the graph, consistent with them often occurring against bottom-feeder teams.

So why do all the graphs start in '09? Didn't Bailiff start in '07?

(12-30-2015 10:43 AM)OptimisticOwl Wrote:  So why do all the graphs start in '09? Didn't Bailiff start in '07?

Lack of time to pull the data. One of my earlier posts said I will update it.

Also, odds are a lot of 2008 data will be tossed out as outliers. A lot of 2011 and 2015 were as part of sanitizing the next set of data (upcoming)

(12-30-2015 10:47 AM)Antarius Wrote:  

(12-30-2015 10:43 AM)OptimisticOwl Wrote:  So why do all the graphs start in '09? Didn't Bailiff start in '07?

Lack of time to pull the data. One of my earlier posts said I will update it.

Also, odds are a lot of 2008 data will be tossed out as outliers. A lot of 2011 and 2015 were as part of sanitizing the next set of data (upcoming)

Curious as to how which data is determined to be tossed.

(12-30-2015 10:50 AM)OptimisticOwl Wrote:  

(12-30-2015 10:47 AM)Antarius Wrote:  

(12-30-2015 10:43 AM)OptimisticOwl Wrote:  So why do all the graphs start in '09? Didn't Bailiff start in '07?

Lack of time to pull the data. One of my earlier posts said I will update it.

Also, odds are a lot of 2008 data will be tossed out as outliers. A lot of 2011 and 2015 were as part of sanitizing the next set of data (upcoming)

Curious as to how which data is determined to be tossed.

stay tuned. the approach will be related to standard deviations though

so far the graphs include all data though, so all graphs and conclusions in this thread are using all data points.

The data is available for review as well at any time.

(12-30-2015 10:50 AM)OptimisticOwl Wrote:  

(12-30-2015 10:47 AM)Antarius Wrote:  

(12-30-2015 10:43 AM)OptimisticOwl Wrote:  So why do all the graphs start in '09? Didn't Bailiff start in '07?

Lack of time to pull the data. One of my earlier posts said I will update it.

Also, odds are a lot of 2008 data will be tossed out as outliers. A lot of 2011 and 2015 were as part of sanitizing the next set of data (upcoming)

Curious as to how which data is determined to be tossed.

2008 had too many good wins to include. So obviously an anomaly.

(12-30-2015 09:18 PM)owl at the moon Wrote:  

(12-30-2015 10:50 AM)OptimisticOwl Wrote:  

(12-30-2015 10:47 AM)Antarius Wrote:  

(12-30-2015 10:43 AM)OptimisticOwl Wrote:  So why do all the graphs start in '09? Didn't Bailiff start in '07?

Lack of time to pull the data. One of my earlier posts said I will update it.

Also, odds are a lot of 2008 data will be tossed out as outliers. A lot of 2011 and 2015 were as part of sanitizing the next set of data (upcoming)

Curious as to how which data is determined to be tossed.

2008 had too many good wins to include. So obviously an anomaly.

Source?

Lets try to keep this thread fact and number based. As mentioned above, all graphs till now include all data without anomalies removed.

The combined record of the 10 teams that the 2008 team defeated was 46-78 (.371).

Antarius - use Massey to go back in history, if you want to tackle the question about the earlier years. He's applied his methodology to essentially all schools, all the way back to 1960. Here's the 1960 graph, and you can go from there: http://www.masseyratings.com/rate.php?s=...1960&sub=0

Unfortunately, not sure how easy it will be to work with this data. May be a lot of manual entry, maybe for one your way back.

(12-30-2015 11:07 PM)WRCisforgotten79 Wrote:  The combined record of the 10 teams that the 2008 team defeated was 46-78 (.371).

Does that mean they won't be outliers, or that they will?