But you see, that's 2011 vs 2014; I'm talking about 1999 vs 2003. And a lot of that owes something to deeper dynamics: in 2003, a lot of "non-target" seats where the ONDP bottomed out around 5% or less in 1999 went up t/w the 10% level, while other single-digit percentages became double-digit percentages. That's where a lot of that overall share increase from 12.6% to 14.7% frittered away to. And on top of that, let me throw in the 2004 federal election, where Jack Layton brought the NDP to 18.1% in Ontario while "only" winning, like its provincial counterpart the previous year, 7 seats--much of that share was due to the party hitting mid-teen shares in hitherto no-hope wastelands like Northumberland and Quinte and what have you. As well as similar increases in the 905.
2014 was different--in a way, more like 1999 than 2003 (or even more so, 2015 than 2011 federally) when it came to the non-target 905 and 416, i.e. the ONDP was pulverized by the urge to vote "strategically", with little incentive to draw that extra bottom-feeder mandate. (And that owed plenty to the respective campaigns of Wynne vs Horwath.)
And also, you refer to the 6% jump in the ONDP *popular* vote in 2014. However, when it comes to *share*, they only gained 1.01% over 2011 (to 23.75%), and that's *with* the added byelection cushioning in Waterloo/London/Niagara/Windsor. And you see, when it comes to psephological geeks, it's all about the *share*, rather than the raw numbers per se...