Good stuff George. It's a great idea to really take a good look at which areas need improvement and then concentrate on those skills. Most people would not analyze their results the way you have here and I'd bet that this will lead to more rapid improvement for you than many other players.
One thing I'll say about the rivet screens is that you probably aren't going deep enough into the game yet to see how nasty they are -- if your PR is 319k I'm guessing your "average" game these days is maybe 250k, maybe somewhere around the end of Level 7 or so. The rivets really don't become a major problem until Level 5, but Level 5 rivets are significantly harder than on previous Levels. If you are commonly playing games up into the Level 10 - 15 range, you'd begin to see how a couple of your men seem to be "stolen" by a seemingly impossible rivet screen. It will be interesting to hear if your opinion of this screen changes as your average game goes deeper.
As for the elevators, a lot of these guys in this community will say things that make it seem like they should be easy, which is likely frustrating to hear. The truth is, these are generally very experienced, kill-screen capable players who are saying this, and they STILL occasionally die here anyways. The fact is, it takes MONTHS of dedicated play to really become comfortable with this screen in the midst of a full game that is going deep, perhaps with a shot at a PR. You absolutely SHOULD set up a save state and practice this way for 10 minutes every day for several weeks -- but there is no substitute for having to deal with this screen on the fly during a good run. I'm confident that most players who have played a serious game every day for a couple of months STILL dread it every time they see the 4 monkeys stacked up, indicating that they will have to deal with yet another elevator screen now. For me, that feeling lasted a long time for games that were going deep.
However, with your overall cag skills and your posted success rate, I have a feeling that you are doing something slightly wrong and/or just haven't developed proper technique yet. You could get that success rate just by accident without actually ever retreating. First (again, with a save state), you should get familiar with how close you can get to the spring that is passing over your head -- most novice players actually get a "poor jump" most of the time because they are timing their approach late enough so that the spring clearly passes over Jumpman's head. This is a mistake. A spring can pretty much pass straight through Jumpman's entire head without a death because of how the hit boxes work for these objects -- you pretty much NEED to have some contact with the spring in order to get a good jump.
The other piece of the technique is spring recognition and responding very quickly with the correct decision. I have a feeling that different players actually do things slightly differently for spring recognition. The technique that I used for a long time focused intently on watching precisely where the spring lands within the yellow box to determine how "long" it is. The trouble is, this is already pretty late in the process and it becomes difficult to make a decision based on this late information. I have now mostly transitioned to using more of a peripheral vision approach which is geared not so much on where the spring is landing, but rather how it is entering the screen. A long spring will enter the screen with high "energy" -- an upward arching motion that just looks threatening, whereas the shorter springs just sort of drop into the screen. This extra millisecond of information makes the decision making much easier, although it's harder to get used to trusting this information since it's more approximate in nature.
Keep us posted on your progress.
