I couldn't resist having a bit a play with the Newsjack numbers (my day job involves lots of spreadsheets, nuff said)
EpEmailsTotal UsedNon-comms usedTotal ChanceNon-Comm chance
153926214.8%3.9%
267527214.0%3.1%
360621163.5%2.7%
451224184.7%3.6%
548127215.6%4.4%
639924216.0%5.3%
Whole Series U6751028615.1%13.1%
Series Recorded67517015025.2%22.9%
EDIT: I don't know how to stop the columns getting squashed here, sorry.
This assumes that no one submitted over the series but the 675 from episode 2 (so the whole series chances would actually be higher as we know this is unlikely to be true). It also assumes that the number of emails includes those from commissioned writers.
Strictly of course, these shouldn't be called the "chance of getting on" as that assumes that getting on is random, which I certainly don't believe. Better to say your work was in the NJ's team's top 5% or whatever that week.
Neither is there some theoretical "best" material each week as if this was a school exam and each piece was graded! The material that best fits the show that week, in the NJ team's opinion, and that gets the best response is what gets used. This will generally be the "best material" in most people's opinions but there will always be room for disagreement around the edges, comedy is not an exact science.
Also they may well cut stuff they really like because it is no longer topical, the story isn't well enough known, or there were, um, legal issues.
To sum up:
1. It's really not an exact science.
2. If you got something, anything on, you were up against a LOT of competition and should be very happy.
3. If you didn't it doesn't mean your work was no good. And all of us can always get better.