Being better is indeed a good way to win more games, but it's not something than happens magicly in the middle of a game. I'm trying to improve my winrate simply by picking better heroes lately and got very good results. I feel like you need to answer the following question with your draft :
- how are we going to win teamfights ?
- how are we going to destroy buildings, especially high ground ?
- how are we going to win the lane phase ?
And I think that the other team was able to answer these question much better than yours. One very big problem in every game is how to initiate teamfight. In my experience, half of the time in pub games the answer is let's watch the other team in the other side of the river for 20 mins and wait for something to happen. And your team doesn't really have a much better option. Tinker does have a dagger, but tinker is a bad hero for initiating fights, he may be able to kill 1 or 2 but he will probably die and dying on a tinker is superbad. I think that your only option for initiating is waiting for them to do a mistake, and it's not a good plan.
Meanwhile they have wraith king which can initiate fights perfectly by just dying, with a sniper being there to secure every possible kill. And the lack of lockdown in your team means that even if a fight goes well, they can all start to run, and with the slow of the drow and the brood you can maybe catch 1 or 2 but it will be very hard to wipe them completly.
Now the other big important thing is buildings, and your team is very bad for attacking high ground. Tinker is great at pushing lanes but is crap against tower. Brood is good for getting a early tower avantage but high ground is much harder, you have to bait the ennemy team out of their base while trying to not loose horribly without the brood.
I think that there are mainly 4 ways to take high ground.
-You can kill the enemy team and then attack buildings. One very good trick for starting this is to ward the enemy jungle and try to catch one of their carries.
-You can use dedicated heroes for taking building, like tiny, sniper, pugna, death prophet.
-You can be tanky enough to not care about the ennemy team.
-You can try to do the rat strategy, but it relies on outmaneuvering the ennemy team. You can relies on enemy mistakes like not carrying a TP when they are attacking your base, but if your ennemies are good enough you need to do good plays like interrupting TPs or convincing the ennemy team to pursue you in half of the map.
The only two plan you can rely are the wipping the ennemy team one and the rat one, but both feel pretty hard to execute and they rely on ennemies mistakes.
One other important thing to consider is your ability to ruin the ennemy plan. Having a not so good plan is ok if you can ruin the ennemy plan. Their teamfight relies on sniper getting kills while their tanky frontline act as a decoy. This is why I think sniper counters like storm or nyx would have been pretty good.
I've been playing a lot of abaddon lately because I think he is a super good answer to a lot of these question on pubs, at least at my level (most of my games appear in high skill in dotabuff). He is one of the best initiation in the game with is ult. It's very hard to screw up when your plan to initiate is getting jumped on in the most horrible way possible. He is super good against building with his passive. He is super good in lane with his shield. And more importantly, he is the closest thing to immortality in dota. I'm starting to be pretty sure he is broken this patch
dotabuff link