"The trend is your friend."

2007-2008 Trend

Free throw Shooting

Free throws - Getting better - not good enough, but unil EDD gets here, it's probably all we can expect.

Assist to Field Goal Ratio

Excellent. Syracuse looks low, but only in comparison.UNC a bit of an outlier, but Rutgers also low - do we have more difficulty play team ball against tough teams?

Turnovers

Turnovers - Syracuse not so good, but Cincinnati looks good, especially with a new lineup. Acceptable at UNC, especially given their ability to force turnovers. Very good against Seton Hall.

Rebounds

Total rebounds fine, even though Tina is struggling, others are picking up. Last five games are distinctly lowere than earleir in the season, maybe limited minutes by Brit?

Defensive Rebounding Efficiency (DRE)

Values over 67% are considered good. Many not so bright spots at SDSU, but the DRE was the best of the year. The DRE at Villanova is a bit misleading, as Harry conceded the rebound to avoid a transition basket. The DRE bounced back from poor Louisville game. Not so hot at UNC. Very solid most of the year, but this is where the rebounding is slipping.

Offensive Rebounding Efficiency

Values over 33% are considered good. While the DRE for Villanova was a special case, the ORE shows just how good they can be. Recent games are very good, except for Seton Hall, surprsiingly low. While UConn shot over 50%, limiting the opportunites, this is a ratio, so it isn't obvious from the stats why it would be so low.

Prior years

2006-2007 Trend

Free throw Shooting

This continues to be a nemesis, but there is a nice, positive trend. Getting better slowly, but a couple more free throws against Tennessee might have changed the outcome of the game.

Assist to Field Goal Ratio

Last year, this stat wasn't very predictive. Same so far this year. Two of the better played games (VA & Purdue) have average ratios. The trend was nicely positive through midyear, then flattened out.

Turnovers

Slightly better than last year, so  far. The recent trend looks good.

Rebounds

Last year, the team average about 37 per game. This year's team is averaging almost ten more per game, but tailed off after Villanova.

2005-2006 Trend

If there's an overall theme for the selected stats below, it has to be "head-scratching". The goal of looking at a stat for a team is to find one that correlates well with winning (or, if we want to be technical, out-performing expectations). One hopes that the stat will look good in wins, and not so good in losses. If that's the case, then it is an attribute worth emphasizing in practice. If one finds such a stat, then it also makes sense to look at the trend over the course of a season, to see if the team is improving. However, if the stat isn't very well correlated with success, it is harder to care about how it changes over the course of the season.

Yet none of the four stats seem to be clear  indicators of success. Free throws are the strongest. Poor free throw shooting stands out in games against UNC and Rutgers. The UNC game was a lost cause - a few more FTs weren't going to change the outcome of the game. The FT percentage at the second Rutgers game was abysmal 38%, but there were only 8 taken. The five misses wouldn't have changed the result. More importantly, if you watched that game, the them of the game wasn't free throw shooting. Ironically, the free throw shooting against Duke was important. We went 5-9. If we go 6-9, we might be talking about our sixth National Championship. That's how important free throw shooting is. I don't suggest for a second this is an insight that needs to be shard with Geno, but 71% is below were a National Championship contender should be.

The other stats are more problematic. Whenever you look at game by game stats, you expect to see some oddities - a very good game where the stat is poor, or vice-versa, but the other stats almost show a perversity. You'd be very hard pressed to pick out the good games by looking at any of the stats.

Free throw Shooting

The second half of the season started out looking very good. Several games in a row near or above 80%. Then results became more erratic, culminating unfortunately, with a subpar game which went to OT.

Assist to Field Goal Ratio

Last year, this didn't seem to be a good indicator of performance, with low values corresponding to good games, and vice-versa.  This year, the conundrum continues.

And still continues. WVA will not go into anyone's list of top ten games by UConn. Maybe not top 100. Maybe not top 500. Yet it was the season high for this stat. What gives?

The entire season followed this theme. Some of it is understandable. If Will is a monster on the offensive boards, and puts back missed shots, who would complain that it hurts this stat. If Ketia or Renee pushes down the court, zipping through the opposition and scoring a layup before the defense has a chance to set up, the crowd will rise to its feet, not bemoan the missed assist. Maybe I need to find a way to count only half-court sets, to eliminate the fast break points, but this still mismeasures the offensive rebound putbacks.

Turnovers

Higher than I'd like, but better than last year so far.

The trend is virtually flat. One of those head-scratching moments. Is the team really not better at the end of the season than the beginning? Part of the answer is that you are measuring against other teams - who are also trying to improve at stripping the ball, outright steals, half-court traps, etc. so a flat trend line could mean you are getting better, but just at the same rate as the opposition.

.

Rebounds

Remarkably predictable through the middle of the season, then a bit more erratic. If anything, a slight downward trend, which isn't good,  but I think some teams place too much weight on rebounds.

2004-2005 Trend

Assist to Field Goal Ratio

Last year, the assist to field goal ratio started out well above 70%, then drifted down, ending slightly below 70%. This year, it started lower, but was trending up, but has been trending down, very erratically. This is one of those stats that may not be a good indication of overall team quality.

Last three games exemplify that observation. All three very fine performances, but very different ratios.

Free Throw Shooting

Last year, the team averaged 68%, a level they should be able to improve. The trend line is finally heading up, but it started from a low base.

Turnovers

Turnovers generally improving, but erratic.

Rebounds

Here's a stat that seems more correlated with results. Five games this year with fewer than 35 rebounds, and we lost all but one. We only had one win with fewer than 35 rebounds (Texas). One loss with more than 35 rebounds, Notre Dame, but then only 36. Only one true anomaly - Michigan State, a loss with 44 rebounds.

Raw rebound numbers suffer from two distortions - a team might have more rebounds simply because they miss more shots and have more opportunities, and the number of potential rebounds will depend on whether the game is a slowdown game or run and gun. See Total rebound efficiency below for a stat adjusting for both of these issues.

Defense

Shooting Percentage Two pointers

Shooting Percentage Three pointers

Total Rebounding Efficiency

An opportunity for a rebound exists when a shot is attempted and not made. (Two exceptions - a missed free throw if the first of two, and blocked shots. This stat adjusts for blocks, but not for free throws.).

2003-2004 Trend

Trends are shown, but the message is mostly an illustration that trends can be misleading.

The trend line is not pretty, but it is barely meaningful. a more important observation than the slope of the trend line is the lack of correlation between "good" games and high assist to field goal ratios.

I posted the above three games ago, and it continues to be true. The Tennessee game was stellar, yet produced the lowest ratio of the year. The Miami win was ugly enough to provoke a "state of the union" speech behind closed doors, yet is one of the highest ratios of the year. The Pittsburgh game returns to a bit of normalcy. An excellent game AND a decent ratio.

The Providence and Villanova games returned to more "normal" relationships. Providence was a blowout, and the second highest ratio of the year. Villanova both a loss and a below average ratio. The Auburn game was one of the best performances of the year, and one of the better ratios.

It is worth noting, that while the trend line continues to be negative, the season ratio, at 69%, is excellent, and slightly ahead of last year's ratio of 68%. We all know how last year turned out.

Don't get too hung up on the downslope of the last four games. They only missed two free throws in each of the last three games. The number of attempts was unusually low. For the first time this year, the team had four consecutive games with free throw shooting better than 70%. That isn't a particularly impressive feat, but this is one of the few areas where the Huskies do not excel, so it is a good sign that even their weaknesses are getting better.

Who can possibly make sense of this? A season low against one of the toughest teams in the country, immediately followed by so many turnovers in the Miami game, that the stat people are looking to see if it is an all-time record. Maybe they focus better in "big" games? So why then did Syracuse return to the season low?

"The only thing we can be sure of: we have no idea how many turnovers will occur in the next game." I wrote this during the wild see-saw period. We've gone to the opposite extreme. The only question now seems to be whether the TO's will be 15 or 16.

.

Conference RPI