Mike Berman's Random Thoughts and Opinions: March 2007

Hillary Clinton Announces National Lawn Care Initiative

Desktop Defense

I got hooked recently for a few nights on Desktop Defense. It may look a bit confusing initially, but it is actually very simple to learn and play, with just a few rules. I taught Pete how to play it in about 2 minutes, and he was able to make it all the way through all 49 of the levels of the "easy" version pretty much by himself. (Warning: it takes about 15 minutes for each full game.)

Here are the stages of Desktop Defense:

1. Beat the game on the Easy level.

2. Start playing the Normal level (because you can't post scores on "Easy").

3. Make it all the way through all 49 levels of Normal level. Post some scores.

4. See how far you can get on the Hard level.

5. Declare decisively that on the Hard level is impossible to get through all 49 levels. Quit the game.

6. Well, maybe if I try... Finally defeat Hard level.

7. Go back and try to complete Normal level without losing a single life.

8. Defeat Hard level without losing a single life.

(Note, I retired after conquering #7 above. I have a feeling that it might be possible to do #8, but no longer have the desire to put in any additional effort to do so.)

Here are my final scores if you want to try to beat them. I say "final" because after my brief de-tox I am not currently planning to ever play this game again.

You can add your scores to our group if you like, which you can see is called "poorneil".

One of the cool things about this game is that the guy just released the current version in March 2007, and 16 million people have already played it (as of March 23). I also saw that it was recently mentioned on Buzzfeed.com.

Sticking With Google

I must candidly say that I am a bit disappointed with the new Kevin Federline search engine. For example, it was not at all helpful in getting me information about Fermat's Last Theorem:

I should look for Fermat's Last Theorem on ebay?

I thought their giving away free t-shirts every hour was classy, but even better is how they include sponsored links both before and after the real results.

No, this just won't do. I am sad to say that from now on, this is the last search engine I will use when I need to find something online*.

* except, of course, for MSN.com

Buzzfeed Pollen Superburst

Yes, below is yours truly getting a shout-out from Buzzfeed. I suggested they cover the "pollen superburst" phenomenon after hearing about it for the first time from a co-worker.

Letter Counting Magic Trick

Here is a magic trick I can do through your computer. It is based on my well-known ability to count the number of letters in words.

1. Pick a number from 1 to 46.
2. Write out the number as a word, with no space or dashes.
3. Count how many letters there are in the written number.
4. Add that number to your number to make a new (higher) number.
5. Write down your new number.
6. Continue this process a few times.

Example: If your first number was 2, your new number would be 5, because "two" has 3 letters, and 2+3=5. The next number would be 9, and so on.

7. Keep going until you reach a number higher than 100.
8. Now close your eyes and think very hard about your current number (the one higher than 100).

When you are ready, click here to see my guess.

So Far, So Good

As a follow-up to my post on how to fill out NCAA brackets, look who is tied for first place among 3.2 million people in the ESPN tournament:

To be tied for first, all you had to do was correctly pick the winners of all 16 games that were played on Thursday. I estimate my chances of winning the entire tournament are now increased from 1 in 3,000,000 to about 1 in 10,000. (In other words, still a snowball's chance in hell).

Tips for filling out NCAA Brackets

This post presents recommendations for filling out brackets for the NCAA college basketball tournament. It is intended for people who do not closely follow college basketball. If you know a lot about college hoops, you should go with your knowledge and these tips will not be as important for you. My qualifications are that I am a mathematics major with expertise in probability theory, and also an avid sports fan.

1. Become familiar with the rules and characteristics of your tournament. These are the things you need to know:
(a) Approximately how many entries will there be
(b) Are there a lot of small prizes, or one big prize?
(c) How is the scoring evaluated? Is it one point for every correct pick? Is there a higher weighting for correct picks in later rounds? Do you get more points for picking a higher number seed to win in an upset?

2. Decide your goals. If you just want to enjoy watching the games, take lower numbered seeds because they are less likely to get eliminated early, so you will have more of a rooting interest later in the tournament. If your goal is to win it all, you will have to pick a lot of upsets. But this means that the odds are that you will be completely eliminated early, possibly by the first Sunday when the Sweet 16 have been determined.

3. The more people who are entering the tournament, the more upsets you will have to pick. Think about it this way - if there were only 2 people in your tourmanent, you should probably take all of the lower seeds (favorites), except for a very small number of upsets where you feel you have a hunch or some insider information. If your tournament had 1 billion people in it, then you should be picking upsets all over the place to uniquely separate your entry from everyone else. As a rule of thumb, make sure you pick a total of about 8 upsets to start. Then you should have 1 additional upset for every 50 people in your tournament, to a maximum of 20 upsets. (There are 63 picks total in an entry). Picking more than 25 upsets will doom you in almost any tournament.

4. If your tournament has a lot of small prizes, pick more of the favorites. If it has one grand prize, pick more upsets - 20 or more. This strategy means your entry is likely to do worse in general, but gives you the best chance to win.

5. If the scoring method is one point for every correct pick, select a medium number of upsets. If there a higher weighting for correct picks in later rounds, such as in the ESPN tournament, pick more of the favorites. If you get more points for picking a lower seed to win in an upset, you should pick more upsets - 20 or more.

6. In the first round, you should strongly favor teams that are seeded 1, 2, 3, 4, 9, 10, 11, and 12. This is especially true in large tournaments that give one point for every correct pick, with minimal weighting. The 1-2-3-4 teams almost never lose, so picking against them in the first round is too risky. The 5-6-7-8 teams are not that much better than the 9-10-11-12 teams, but a lot of people favor them. So picking the 9-10-11-12 teams gives you separation from the field without significant additional risk. You should also know that 12 seeds have done surprisingly well against 5 seeds in the last 10 years. But I don't think this trend has any real basis, or will continue in the future.

7. Don't go crazy with the upsets in later rounds. Hoping for multiple upsets in later rounds makes your bracket too risky. The only exception is very large tournaments, or tournaments that give heavy weighting to selecting upsets. At least 2 of your final 4 teams should be quality teams that are a 4 seed or lower (better).

8. If it's a free tournament, enter as many times as you are permitted. If you are submitting multiple entries, make your picks in different entries radically different from one another. I would recommend not even having any of the same teams in the final four. This will diversify your risk, so you do not get multiple entries eliminated by one surprise upset. It might feel painful to pick against a team that you think is "a lock", but these upsets happen all the time. When George Mason beat UConn in 2006, many people who had UConn in the final four of all of their entries were completely eliminated.

9. If you have some familiarity with the teams, you should favor teams who have performed better late in the season. Even though this is factored to some extent into the seedings, teams with better recent records tend to do better in the tournament than even their seeding would predict.

10. If you are doing the women's tournament, pick fewer upsets than you would for the men's tournament.

11. Remember that there is a very high amount of randomness and luck. So the best thing to do is make your bracket relatively unique from the others who are entering and then hope for things to bounce your way. All of the brilliant "experts" who pontificate prior to the tournament are always heard joking about "tossing their brackets in the trash" on the first Monday morning.

Are You Smarter Than a Fifth Grader

Some random observations about the new Fox game show "Are You Smarter Than a Fifth Grader":

I only watch it because my kids like it. I swear.
It is a shameless ripoff of Who Wants To Be a Millionaire. But they give away less money because there are a bunch of cute kids in the mix.
The adult contestants are college graduates who generally do very poorly. I estimate that they get the questions right about half the time. The first contestant we saw, on the premiere, got zero questions correct out of four.
The questions claim to be elementary school questions, but they are generally much harder than what most people would classify as that level question.
Despite the difficulty of the questions, I have watched 4-5 episodes and have NEVER YET SEEN A SINGLE KID GET ANY QUESTION WRONG. I am not big on conspiracy theories, so I am inclined to believe that these are just hyper-intelligent kids (are they supposed to be normal or geniusus? I don't recall if they tell us that.) But it seems to me that, even though these are facts that you tend to forget over the years, the odds of the adult college graduates getting 50% and the kids getting 100% would be about 900 trillion to 1. So I think something is going on. Maybe they are giving the kids study sheets, or something. But this makes the show impossible for me to watch, because it defies credulity and may be lacking in integrity (if the conspiracy theory is true).
Example: On the first or second episode, they asked who was the first U.S. President to be impeached. All 5 of the 10 year old kids correctly wrote Andrew Johnson (it was not multiple choice). Give me a break.
This show seems to take a perverse joy in adults being stupid. It's the only game show I can recall where everyone laughs about what an idiot the contestant is. In fact, they require every contestant to leave the show by looking in the camera to say that they are dumber than a 10 year old.
Jeff Foxworthy is horrible.

Tron!

I was quite pleased to find this cool online version of Tron light cycles the other day. First, it reminded me of writing the very same program for the Apple //e back in 1985 (which I still have working in my basement). Second, it reminded me of one of my favorite video games. Third, it presented quite a challenge. Sometime you should try it to see how you can do.