Why 538 is wrong, and why Clinton is in better shape than Kerry was

538 wrote:

Based on the polls, we think the model is setting those odds about right. The race is a long way from being a toss-up, but a 3 or 4 percentage point lead heading into the conventions isn’t all that reliable, either. While Obama won twice with pre-convention leads of about that margin, John Kerry went into his convention with a lead of about 3 percentage points in 2004, but lost to George W. Bush. And in 2000, Bush had about a 4-point lead on the eve of the conventions, but lost the popular vote to Al Gore. (Bush won the Electoral College, of course.)

They then post a summary convention bounces. I wrote a much longer history of convention bounces (which goes further back than 538, and is frankly better) at Bleeding Heartland.

Here is the problem: their analysis is very wrong.  Very VERY wrong, and shows a lack of understanding around what causes movement around conventions.  

In 2012 here I wrote two diaries about the Republicans.  In the first diary I wrote this:

What the Kerry experience strongly suggests is that naming your VP selection early may substantially effect the bounce you get from a convention.   For non-incumbents, it appears that the VP choice has an enormous impact on the public's perception of the candidate, and is responsible for much of the bounce that they receive.  

In that diary I posted the following chart to make this point. So at this point in 2004 Kerry had already named his VP, and as a result was predictably not likely to get much of a bounce out of the convention. 

As a result the comparison to Kerry at this point IS COMPLETELY INVALID.  

In 2012 I wrote a second diary, noting that Romney really had not received a bounce out of the convention.  I noted at the time that like Kerry in 2004, Romney did appear to get some bounce out of naming Ryan.  But just in 2004, this bounce faded.

This table summarizes the effect prior to 2012.

The point here is that much of the analysis about bounces is fundamentally flawed.  Really 538 has shown they don’t understand how convention bounces work.

Hillary Clinton has not named her VP.  She will get a bounce out of that and the Convention.  At this point in the process in 2004 Kerry had named his VP, and could expect no bounce out of the convention.

538 is completely wrong.