Kurtosis vs. tails vs. t-distibution

According to the curriculum (and wikipedia) higher kurtosis means higher peaks and fatter tails, while lower kurtosis means just the opposite. T-distribution is by definition leptokurtic, but the higher the degrees of freedom - the thinner the tails. Keeping the characteristics of kurtosis in mind, what I would also expect from t-distribution is that it would also get less picked as we increase the df, but in fact it does the opposite: peakedness increases. Perhaps, I have misunderstood one of the definition/properties, but I can’t figure out which on. I would appreciate some insight.

Eh? The t-distribution has LESS peaked than the normal distribution and has fatter tails. Leptokurtic, on the other hand, is more peaked and has fatter tails than the normal.

t distribution is leptokurtic. Leptokurtic mean more peaked (which means the distribution close to the mean does not have as many elements, it is thinner) It could be higher than normal distribution’s peak or lower. Just that the distribution is thinner… the tails are fatter…

99 cannon sloop, according to http://en.wikipedia.org/wiki/T-distribution T-distribution is always leptokurtic (as long as df > 4, ex. kurtosis is >0). anish, makes sense now. So, apparently, I was confusing peakedness with the height of the peak. Thanks a lot!