Re: NAQT ICT Picks and Heartland Clarification

• From: jagluski <no_reply_at_yahoogroups.com>
• Date: Thu, 13 Feb 2003 06:00:34 -0000
• Message-ID: <b2fca2+m749_at_...>

Just to clarify the Heartland schedule fiasco, the teams played off 
the original schedule.  So, in round 1, 1v2, 3v4, 5v6 etc.  4 and 5 
didn't show, so 3 played 6, even though they were scheduled to play 
later.  Thus, I have no idea how you would even factor out those 
matches, since one match occurred each round that was a duplicate.

Joel Gluskin
Wash U Academic Team, President




--- In quizbowl_at_yahoogroups.com, jpaikman <no_reply_at_y...> wrote:
> > Jason, it's an interesting formula... but I think your Heartland 
> > Sectional calculations are a bit flawed.  For one thing, 
thoseteams 
> > should not get credit for the forfeit win vs. Wichita State (it 
> > should be completely taken out of the equation).  Also, wasn't 
the 
> > Heartland sectional the one where they played random extra games 
in 
> > addition to the round robin, because the 13th team didn't show 
up?  
> > That being the case, a strength of schedule adjustment may be 
> > needed.
> 
> Point taken.  It's impossible to predict the effect of random games 
> in addition to the round robin, and I'd think that NAQT will also 
> have that same problem.  None of us have access to standings before 
> those two random matches.
> 
> > Also, I think that the winning percentage factor may be a 
bithigh.  
> > Whether or not it _should_ be that high is another issue, but as 
a 
> > long-time MAQT fan, I remember a 9-4 Maryland team that did not 
> > advance to the ICT (while a 6-7 team from another school did).  
In 
> > other words, I'm not so sure that Truman State A, for example, is 
a 
> > shoo-in to make the field.
> 
> I originally didn't want to include an adjustment for winning 
> percentage, and just base the ranking on bonus conversion and power 
> tossups.  These are the only two equalities amongst all sectionals -
 
> in fact, disregarding those two random matches above, using an 
> opponent-blind statistic such as bonus conversion and power tossups 
> gives out the same teams on top.  In general, better teams win.
> 
> In any case, winning percentage for Division II teams counts for 
> about 50% [definitely high], but counts for about 33% in Division 
I, 
> which is more appropriate in the NAQT scheme.
> 
> Moreover, under recalculation where the winning percentage 
adjustment 
> is cut in half, the top 24 teams still remain the same, with 
changes 
> in ordering.
> 
> ***However, the goal of this experiment is to predict finish at the 
> ICT.***
> 
> > Finally, this year NAQT is taking 24 four-year and 8 two-year 
teams 
> > for the ICT, right?
> 
> I'm not aware of this.  I'm sort of out of the loop.  If that's 
true, 
> then its obvious what adjustments have to be made.
> 
> ===================
> 
> MIT B
> Harvard B *
> Columbia
> Brandeis Parasite *
> Williams B
> 
> Florida C *
> Georgia B
> Florida B
> [UTC may use its autobid here.]
> 
> North Carolina A
> Princeton B
> Virginia
> Swarthmore A *
> 
> Michigan D *
> 
> Rice A *
> Tulane A
> 
> Chicago *
> Carleton B
> 
> Arkansas B
> Iowa State B *
> WashU B
> Oklahoma C
> 
> Cal-Berkeley *
> Caltech
> 
> British Columbia B *
> 
> ===================
> 
> Thank you for your criticism.
> 
> Jason Paik
> QB Ronin (TM)