Add Tricky Level Punisher

[Nikoleta-v3]

I picked up #1132 which implements a strategy called Tricky Level Punisher.

The pr was failing due a division by zero error. Moreover, I had to tweak the tests of MetaMajorityLongMemory because now TrickyLevelPunisher is in the strategy's team.

>> import axelrod as axl

>> player = axl.MetaMajorityLongMemory()
>> tricky = axl.TrickyLevelPunisher()
>> tricky in player.team
True

Note that this pr will fail due to #1176, I will rebase once #1177 is merged.

[marcharper]

Rather than update the meta player tests each time, which amounts to a "change detector test" at the moment and can be a pain to update, we might consider simply having a test for the number of team members.

[Nikoleta-v3]

I am happy to fix the tests for the meta players in a different pr if you all agree.

[drvinceknight]

Rather than update the meta player tests each time, which amounts to a "change detector test" at the moment and can be a pain to update, we might consider simply having a test for the number of team members.

Yeah great call @marcharper, perhaps not just the number but a test of the makeup of the tea (specific members?) also good to have a test that matches complete too. (@Nikoleta-v3 if you could handle this that'd be great 👍)

#1177 is merged so you should be ok to rebase.

[drvinceknight]

Thanks @Nikoleta-v3 merging this in 👍