You're probably commenting on this statement: "The average minutes per game for all 12 matches was 6.8. The elapsed time was measured from the lag until the winning ball was made, so it includes time for racking, player timeouts, and commercial timeouts (last match)."
No typo there. I certainly am not intending to say that every game is started with a lag. The second sentence of my statement relates to two points in time -- when the balls are struck for the lag for a match and when the match-winning ball is pocketed (or the cue ball stops rolling). And the time for the lag (and for racking the balls for the first game if they were not racked prior to the lag) is, to me, certainly part of the elapsed time for the match. So it is included in the elapsed time I compute for each match, even though it s-l-g-h-t-l-y lengthens the elapsed time. I would guess that the lag itself rarely adds anything to the average minutes per game calculated to the nearest tenth of a minute, but if the balls are racked for the first game after the lag, that can affect the average slightly, just as racking for all of the other games does.
Now, it would be possible to calculate just the actual play time in a match, by eliminating the time spent on the lag, the racking for each game, the player timeouts, any commercial timeouts, any referee-called timeouts, etc. But that's not something I want to do. Many years ago, when I first started keeping track of, and then reporting, match lengths and average minutes per game, I thought that information might be of some interest to people planning tournaments. But I don't know whether anyone ever used my information for that purpose.