In making his argument in support of the Republican "electoral college lock" of recent years, David Broder {op-ed, Nov. 22} uses the last seven presidential elections, of which the Republicans have won four and the Democrats three -- in and of itself a "lock" without great practical effect.

Of these seven elections, Republicans have won in landslides three times and Democrats once; three have been close contests. If one agrees that the 1988 results are more likely to track the close elections than the landslides, then lumping the landslide data with that of the close contests skews the seven-election data for 1988 predictive purposes.

The three close elections held since 1960 had certain characteristics in common, none shared by the four landslides:

No elected incumbent at the top of either ticket.

One party coming off eight years in office.

A party-oriented, nonideological candidate at the top of each ticket.

The first two of these characteristics describe 1988, and the third is likely to.

The three elections, of course, are the close contests of 1960, 1968 and 1976 -- of which the Democrats won two and the Republicans one. So, depending on your analysis, the Democrats are either down 3-4 or ahead 2-1. Cubs fan Broder knows that with a one-run margin, anything can happen. F. T. MERRILL JR. Washington