The first day of the three-day California State Fair wine competition is over, and my mouth feels - and looks, I fear - like a raisin. The astringency of 96 young zinfandels will do that to your gums, lips and tongue. (By young, I mean all these zinfandels were from the 2004 harvest; most of them are just being released for public consumption or won't be released for another month or two.)

The State Fair's 72 commercial wine judges sit four or five persons to a panel. The workload for each panel is about evenly distributed. My fellow panelists are Rod Byers, sales director for Nevada City Winery in Nevada City; Carol Shelton, owner and winemaker of Carol Shelton Wines in Santa Rosa; and Rosina Tinari Wilson, senior editor of Wine X Magazine of Santa Rosa.

At big wine competitions, larger classes of wine - say 30 or more wines of a type - customarily are delivered to panels in flights of 10 or 12 each. Sometimes panels award medals flight by flight, sometimes they vote to eliminate or retain each wine, then reconsider the retained wines for medals the next day.

While the State Fair is a large competition - 3,000 or so wines - it is perhaps unique in the country for its use of the "Peterson method" in evaluating wines. Named after its American instigator, longtime California winemaker Richard Peterson, this approach to wine evaluation calls for panels assigned large classes of wine to be served huge flights, generally 30 to 40 wines each.

Judges are to deal with such large groups by first sniffing each wine, and on the basis of smell alone set it aside as a potential gold-medal wine, silver-medal wine, bronze-medal wine, or eliminate it from further consideration altogether. Thus, after going through a large group of wines each judge will have before him or her three or four smaller clusters of wine. Then they start tasting, beginning with the gold-medal candidates. They jot notes, sometimes move wines from group to group as they alter their impressions, and eventually arrive at a confident assignment of awards.

The principle behind the Peterson method is that if judges rely more on nose than palate they won't get as fatigued, and their assessments will be more consistent. I don't know whether this has been validated - Peterson often judges at the State Fair, but not this year - but I do sense that after tasting nearly 100 zinfandels with this approach I'm not as beat as I generally am after judging the same number of wines with the traditional approach. I won't speculate on whether results are more consistent or meaningful, but the proportion of medals awarded does seem to increase when the Peterson method is used.

At any rate, up to this point in the competition judges have been working individually and quietly. Then they gather together and recite their scores to a clerk, who tallies the votes and comes to a consensus medal for each wine, or no award. When judges differ widely on their appraisal of a particular wine, they talk out their differences and generally eventually concur on what sort of medal it is to get, if any.

Our panel tended to differ more than we agreed, but generally not widely, though there were times when one judge would think a particular zinfandel should get a gold and another judge would think it shouldn't get any medal at all.

How to account for such differences? Different sensitivities, for one. Judges have different thresholds for what they consider acceptable levels of ripeness, alcohol, tannin, oak and so forth in a wine. Stylistic preferences also factor into the equation. Some judges may like their zinfandels light and frisky, others dense and brooding. Judging is a matter of giving and taking, learning from each other, and keeping an open mind.

I'll try to keep all that in mind when judging resumes Saturday morning. We have 81 more zinfandels awaiting us, plus six malbecs.