The problem is still that you answered the question with only some outcomes on the table. You can only usefully answer the question if you apply the same logic without knowing which party is ahead to all ridings. ...
I do not think this process seen in hindsight is a good way to look at it unless you recognize that people did not know which party was ahead and so the advice would go across all seats where this is not known.
Otherwise you only get the answer for how those with a crystal ball ought to apply their vote strategically.
There is no way to answer any question without considering "only some outcomes". That is the nature of analysis. No one can consider every outcome in a single analysis (let alone something appropriate in length to a message board) -- there are far too many outcomes to postulate.
Again, I was not advocating that people vote strategically. If your position is that, one week afte the election is over, it is not permissible to ask what would have happened if people had voted in a particular different way, then I have to disagree that that is not an acceptable thing to wonder about. In fact, I think that the best time to wonder about "what ifs" is shortly after an election has happened, not in the run-up to the next election. I cannot imagine that anything in my answer would have any impact on anyone's future plans to vote strategically, in an election 4 years from now.
In the days after an election, NorthReport asked what would the results have been a few days ago if X had happened. I did the number crunching to give one version of X happening. Elections results always are seen "in hindsight" -- election predictions are made in foresight (of wildly varying degrees of accuracy, mostly imperfect). Are you suggesting that no one should ever look back at an election's results and wonder "what if" or just ask a question?
No - you are missing the point. You are looking at the math of what ifs on seats the NDP were a lose second to the Conservatives. You ahve to apply this across the board as people in advance cannot make that kind of selection. If you apply it across the board the NDP gain some seats but the Liberals lose some to the Conservatives as well. This is the issue we have raised on behalf of the NDP for years.
So I am not saying you cannot do a what if excercise but you cannot apply this to only a few seats selected based on knowledge only the election result can give -- and get anything worthwhile.
In any of these strategic voting attempts, if you are trying to drive to one party then oyu have to accept the cost of the errors when that party you are driving to is not the leading party and the planners misjudged.
In my own riding I thought the Liberals were running third and it would be close between the NDP and the Conservatives. The polls said so and the NDP had a good candidate. If there had been a strategic voting swing here then the Conservatives would have won.
Selecting only the onese won by the conservatives to apply the math to is not going to give you a net answer of what would have happened.
On top of this when you introduce the concpet there will be mistakes made in both directions and seats potentially lost. Voters on their own cannot do this.
This is apart from the issue that the NDP is as distinct from the Liberals as the Conservatives are sopredicting what would happen to the Liberal vote is not easy.