Coming off a difficult defensive inning in the bottom of the 9th in Cleveland, the Jays provided the faithful with a wonderful effort on Labour Day versus the Yankees. A stellar defensive play by Reed Johnson, throwing out Nick Johnson at 2nd base for a double-oops-single, snuffed out a potentially big top of the first for the Yankees. Roy Halladay was nearly flawless the rest of the way, and the Jays got offense from up and down the lineup to win going away.

That kind of baseball is why many of us thought of them as candidates to win 90 games this year earlier in the season. One problem has been the occasional stretches of poor, sometimes abysmal, defensive play. Contenders don't often beat themselves and the Jays' defense and bullpen has done too much of it.

Although the Jays now have several good defensive players, many quality hitters and a decent pitching staff (adjusting for park), they remain only an average ballclub. Their defence has been sub-par most of the year. This is discouraging, since the emphasis this spring was on playing fundamentally sound baseball. And when we think of fundamentals, we naturally think of execution on the defensive side of the ball.

As the following tables indicate, the Blue Jays have played arguably the worst fundamental baseball on defence in the American League this year. The three categories of performance - wild pitches and passed balls, opponent base stealing, and errors - are all areas we would expect a team focused on defensive execution to do well in. The problem with the blueprint is that most inexperienced players lack consistency in the field, even if they possess better defensive tools than most veterans.

The first table charts the performance of each American League team in three "fundamental" defensive categories. Wild pitches and passed balls are divided by the number of pitches thrown with men on base; errors are divided by number of balls in play (eliminating the putouts on strikeouts as well as extra - unnecessary - assists).

According to play-by-play analysis, the value of a passed ball or wild pitch over the last 4 years is approximately 0.285 runs (see the last line of this table). Each wild pitch or passed ball above or below what a league average team would commit given the same number of opportunities is converted to runs by multiplying by 0.285. The results are in column 2 of the following table.

Similarly, a stolen base has a positive value of 0.195 runs and a caught stealing is worth -0.456 runs to the offence. The implied break-even point is about 70%. Turning that around (i.e. from the perspective of the defence), we can determine how many runs relative to league average the catchers/pitchers have gained or lost on opposition stolen base attempts. The results are in column 5.

Finally, the cost of an error depends on what type it is. Errors which allow the runner to reach safely cost the defence 0.546 runs on average. Errors that allow a baserunner one or more extra bases are assumed to cost 0.285 runs (the same as WP/PB). I don't know how many of each type actually occured, so I assumed a 50/50 split, which implies a cost to the defence of 0.415 runs for each error. Again, the number of errors above or below league average given the number of opportunities was multiplied by 0.415 to determine net runs on errors (column 7).

The Blue Jays have been bad in all three categories this season - no better than 12th in any single category. No other team is in the bottom half of all three categories.

By my estimation, the Jays are weak defensively at three positions - third base, shortstop (when Bordick doesn't play) and catcher. Yet, this team isn't far away from fielding a good defence. Carlos Delgado has performed admirably at first base. Orlando Hudson was error-prone earlier in the year but has since settled down and is athletic enough to handle second. Reed Johnson has greatly improved the outfield defence with his accurate throwing and Vernon Wells has great range and is improving his routes to the ball. The hope is that Kevin Cash will limit the opposition running game, but he's at least a year away from being able to do that. Wilson and Myers are probably a little below average in the throwing department. I like the way the middle infielders turn the double play.

The pitchers need to help the catchers out. Kelvim Escobar, in particular, seems to forget about baserunners (18 SBA in 20 tries this year). I hesitate to suggest any changes to Roy Halladay's game, but baserunners are 23 of 26 off him in steal attempts in 2003. With two starters like Kelvim and Roy, an effort ought to be made to find other starters who are adept at cutting off the running game. Hendrickson, despite being slow to unwind, could be one of those starters - if he comes up with a deceptive pick-off move.