There are many things to be said of Tony Blair, and by now I imagine all of them have been said repeatedly.
His government deliberately delayed progress on LGB rights, and was the only government to ever legislate away the hard-won human rights of transgender people. It gave the NHS a short-term fix that was consciously designed to bankrupt the organisation in the long term, relying on Labour being out of power by the time the chickens came home to roost. It brought in a - most welcome - minimum wage that the Liberals were attacked for critiquing by the unions and then a decade later the unions started attacking it using the self-same lines as the Liberals had used. It delivered partial devolution of power but kept in the hands of the PM the valuable power to appoint peers as a way of bribing people to toe the party line. There was a spending boom off the back of a bubble that the government totally screwed up and nearly wrecked the economy. There was then another spending boom off the back of a different bubble that the government totally screwed up and really wrecked the economy. Iraq. The trans tax. And so on. You know the score.
However, what Blair did that no other Labour leader has managed in my life, was to properly win. As a general rule of thumb no matter how good or worthy or wicked or cruel your plans, no matter how well you have spun them this way or that, if you don't get in then it is unlikely to come to pass. Very occasionally you get your plans into action regardless: the NHS was implemented, albeit somewhat botched, despite the party behind it not getting into power in 1945; phase 2 of the Bedroom Tax was rolled out similarly in 2010 despite its architects having been shown the door.
What Blair realised - and maybe it needed losing four times in a row to give the vision painful clarity - was that for all the theoretical and philosophical stuff, in practice there are seats Labour can win from the Tories, and there are seats the Liberals can win from the Tories. There is a small amount of drift, perhaps 1% a year, in each seat on this but the general lesson is that there are places where for example the least Labour can get is 5% and the most is 30%: either way they lose, but a couple of thousand people's decisions about whether to vote Labour decides whether the seat goes Conservative or not.
In 1997 Labour and its activists engaged in a pincer movement with their counterparts in the Liberals. To be broad brush and so slightly misleading about it, Blue-Gold seats got hammered by the Liberals and Labour stayed quiet or went elsewhere; in Blue-Reds the Liberals sat on their hands while Labour went to work. Tories found they were fighting on two fronts against an enemy that has declared more-or-less a truce on its own internal battle front. Result: a stonking Labour majority and a big Liberal bloc with the Tories out of power for a decade.
Contrast to 2010, 2015, 2017 and - on the YouGov poll last night though it ain't over etc etc - 2019. Labour keep setting off on a battle plan involving taking out two enemies at once, despite having seen time and again before how that is a strategy entitled How To Get Second Place But Not Win.
General Melchett would be proud of them.