More thoughts on the 200-year test

(This of course refers to the previous post about the question, Will any pop music made since 1950 last 200 years?)

*Pop music with lyrics, as opposed to, say, instrumental classical music, depends not just on quality but also on context. It captures listeners' lives and attitudes -- and lives and attitudes can shift dramatically in 200 years. For a song to last two centuries, it must have a quality that's so high it makes up for a near-zero in the context category, a hurdle that was lower in centuries past. Therefore, Murray might be wrong that "accomplishment" in music is declining; music is perhaps just becoming more context-dependent, with a new song for each situation.

*Partially for this reason, most people stick with the music they heard growing up. Everyone who grew up between 1950 and today will be dead by 2200.

*Some people are able to pass these tastes on to their kids, but this weakens with each generation. The question: If my dad listens to a lot of Beatles and I listen to a little Beatles, will my great-great-whatever grandchildren listen to any Beatles at all? In other words, will interest in the best music from the last 50 years even off before reaching zero?

*If something does last 200 years, we shouldn't assume it will be something that was fully appreciated in its time. Bach wasn't. Maybe some independent-label band from the mid-1980s will finally get the respect it's due in 2185. Or maybe the Bach of Pop died last year, and his wife is discovering his home recordings in the attic as you read this.

Lord no

Charles Murray once proposed that the number of artistic works since 1950 that would survive 200 years was close to zero. He suggested that, each time you think of a contender, you have to ask yourself, "Seriously?"

Steve Sailer points to a list of possible exceptions from The Man Who is Thursday. Pop music is my only real area of expertise here -- and the list includes Nirvana. Seriously?

Nirvana added absolutely nothing to pop music. Did they write some catchy songs? Sure. But they didn't break out of the pop mold, and their kickstarting the "grunge" movement was sheer luck. Other bands with similar sounds had been around for years. If you had to pick a grunge band whose work would endure, and I'm not saying any of theirs will, you'd have to go with Alice in Chains (much better vocal harmonies) or maybe Soundgarden (cool use of time signature shifts).

Excepting drummer Dave Grohl, Nirvana members couldn't play their instruments. Allowing that they had some decent vocal harmonies here and there, their compositions lacked innovation and depth. Mindlessly catchy songs come and go. They were an OK band, but they've been overrated since their breakthrough.

Through their toned-down image and dirty sound, they offered an escape from hair metal -- much as Green Day's happier, brighter sound offered an escape from Nirvana a few years later -- and without that lead-in their appeal disappears. People who lived through that era (me included) will always like hearing "Smells Like Teen Spirit" on the radio, but in 200 years we'll all be dead, and the thrill will be gone.

Mentioning Nirvana in the same breath as The Beatles, Bob Marley and Bob Dylan? Seriously?

AK-47 facts and fiction

I have to say this every time there's a media storm around a multiple-victim public shooting, but the gun used in Nebraska was not a real AK-47, as some media outlets have been implying. An AK-47 is an automatic weapon, meaning when you hold down the trigger it fires continuously and rapidly (in other words, a machine gun). Civilian versions, far more common in the U.S., are semiautomatic, meaning they fire once for each pull of the trigger.

In fact, the gun wasn't even a civilian AK-47. It was an SKS, a semiautomatic rifle that preceded the AK-47. The LA Times even calls it both in the same article.

Also, John Lott points out that the incident occurred in, not surprisingly, a "gun-free zone" where people can't defend themselves.

UPDATE: The LA Times has removed the SKS reference from its article, leaving only the AK-47. Different outlets are calling it different things, so I'm no longer sure it was an SKS rather than an AK-47 -- I suspect as much, since if half the media has it wrong, it's probably the half that's using the menacing-sounding, well-known name. Regardless, my main point stands: There is no evidence from any report that the gun was a true AK-47, even if it was an AK-47 in name. It was semiautomatic, making it no more or less lethal than a hunting rifle.

UPDATE II: It appears the LA Times got it right; it was an AK-47, "which [investigators] initially had identified as an SKS assault rifle." I'm really of two minds about a few grafs from the story:

The AK-47 is a semiautomatic assault weapon that generally is used by the military or collected by gun enthusiasts.

It is not accurate for long-range shooting, which makes the gun impractical as a hunting rifle. The automatic version is popular for military use because it is dependable and easy to manufacture.

On the plus side, it makes it clear that the gun was semiautomatic (as opposed to the military version), but it doesn't define the term. Also, it doesn't mention that it uses a bullet size comparable to that of a hunting rifle -- same bullets + same firing speed + less accuracy = less dangerous, so that's important information. It seems like the writer cared more about the "you can't use this for hunting!!!!!" argument than the "this gun is no more dangerous than regular hunting guns" fact.

Abstinence debate, meet scientific method

I won't claim to have read every study of abstinence education, but to my knowledge it's never been shown to reduce sex-related problems like STDs and pregnancy.

Still, people need to cut it out with the garbage arguments. The teen birth rate rose in 2006, and The New York Times takes seriously the question of whether abstinence-teaching caused it. Lawyers, Guns and Money thinks so. Matthew Yglesias piles on.

Let's take a closer look. Federal funding has been rising, sometimes slowly and sometimes by leaps and bounds, since at least 1997. 2006 marks the first time since 1991 that the birth rate rose -- in other words, it's the first year in at least a decade that the raw data has run counter to the "abstinence ed decreases the birth rate" thesis.

Experiments (taking three comparable groups and putting one through abstinence education, another through regular sex ed and the last through no program) and regression analyses (looking at what education kids are getting and whether they get pregnant, and controlling for variables like income, race, etc.) are far better ways to answer this question, so of course this doesn't prove abstinence education works. But the left needs to stop overreacting whenever data even slightly supports its ideas.

And as an aside, the increase is small; judging by the graph, the birthrate is about as high as it was in 2002-3 or so. In addition, blacks seemed to demonstrate the most dramatic 2006 spike, and they experienced an equally dramatic 2005 valley -- once a given statistic gets very high or very low, it's hard to drive it further in that direction, so a minor setback shouldn't be taken as evidence of failure. Finally, in some small part Hispanic population growth is causing this, as they have double the average teen birth rate and are both immigrating and reproducing rapidly.

UPDATE: At Reason, Kerry Howley weighs in helpfully: "...kids are not as malleable as supporters of comprehensive sex ed policies make them out to be. The available evidence suggests that abstinence-only programs have no impact whatsoever. Kids might as well spend the 40 minutes staring at a brick wall. ... You could argue that schools ought to convey accurate rather than inaccurate information about the subject. I would agree with you. But as far as I know, there is no solid evidence that 'comprehensive' sex ed--the relevant alternative--has any impact on sexual behavior either. That's the conclusion of UC Berkeley sociologist Kristin Luker's extremely thorough book on the subject, where she explains why she can't find a single study robust enough to back."

Survival of the fittest and prediction

Yesterday I attended the Robert Taft Club's "Darwin, Genetics, and Conservatism" event, featuring Charles Murray, John Derbyshire, Ron Bailey and Tom Bethell. (Full disclosure: I've been to many events and enjoyed them, so I joined the club.) Bethell made a point that Ann Coulter once did: In "survival of the fittest," the fittest are defined as those who survive. Thus, one can make up an explanation for anything, and it's not useful.

If a certain species is aggressive, it must have faced situations where it needed to be aggressive. All you have to do is find a situation where that would be the case (say, fighting a predator or capturing prey) and voila, you've explained it. You don't have to prove that the value of aggression actually outweighed the value of passivity (getting along with others, not being detected by predators in the first place) -- natural selection is assumed, so the fact that aggression "won" is proof enough.

However, I think it's plainly obvious that as a concept, "survival of the fittest" has a lot of value. If a certain trait leads to premature death in an organism, that organism won't be able to pass its genes on. If you introduce a predator that runs fast enough to kill off many lizards, the remaining lizards will have longer legs (on average), and the next generation will likewise have longer legs than the previous one did.

Thus, survival of the fittest is more testable when it predicts the future as opposed to explaining the past. With the past we already have the outcome -- Species A is aggressive -- so we can only assume that natural selection took place and try to figure out how. It serves more as an analytical framework than a testable hypothesis, unless you happen to have ridiculously detailed information about a species' environment through history, allowing you to weigh the advantages of two different traits.

At this point there's no doubt that some aspects of evolutionary theory are true. DNA tests show common descent, and we have seen genetic mutations and "survival of the fittest" occur. The big question is whether these concepts can explain all of evolution, or if there's another mechanism involved. Positing some designer, as Intelligent Design folks do, is a bit of a cop-out (we don't know how it happened, so let's say there's this really smart thing that directed it), but I wouldn't be surprised if scientists discovered some process that complements the established genetic mutation/natural selection theory.

If that happened, it would prove both Darwinists (mutation and selection explain everything) and ID people (there is no way this could have happened without a designer) wrong.

Too smart for your own good

Tyler Cowen points to this article on how to raise bright children:

Our society worships talent, and many people assume that possessing superior intelligence or ability—along with confidence in that ability—is a recipe for success. In fact, however, more than 30 years of scientific investigation suggests that an overemphasis on intellect or talent leaves people vulnerable to failure, fearful of challenges and unwilling to remedy their shortcomings.

The result plays out in children ... who coast through the early grades under the dangerous notion that no-effort academic achievement defines them as smart or gifted. Such children hold an implicit belief that intelligence is innate and fixed, making striving to learn seem far less important than being (or looking) smart. This belief also makes them see challenges, mistakes and even the need to exert effort as threats to their ego rather than as opportunities to improve. And it causes them to lose confidence and motivation when the work is no longer easy for them.

In other words, we need to knock smart kids down a peg.

This is true, but not all that groundbreaking. It relates to the psychology concept "locus of control" -- someone who has an internal locus of control believes his behavior affects his life; someone with an external locus of control believes things just happen to him. If you tell a kid he's doing well in school because he's just plain super, he'll take later struggles to mean his super-ness has worn off. If instead you encourage a kid to work hard, and present him with material that challenges him, he'll learn that his effort, as opposed to just his biological endowment, is connected to his performance.

Charles Murray once noted another reason smart children should be taught lower self-esteem:

[C]hildren who know they are smarter than the other kids tend, in a most human reaction, to think of themselves as superior to them. Because giftedness is not to be talked about, no one tells high-IQ children explicitly, forcefully and repeatedly that their intellectual talent is a gift. That they are not superior human beings, but lucky ones.

A little humility goes a long way.

Whoopi Goldberg, Ron Paul and multiple-abortion statistics

Reason magazine highlights a clip of Ron Paul debating abortion on The View. Whoopi Goldberg points out how she really doesn't like abortion, but it should still be legal. The logic: No one wants to have one, so when they choose to, it's the result of careful consideration.

One way to test this theory is by looking at multiple abortions -- if a woman sees abortion as something to avoid, and then a decision to (ahem) labor over, one wouldn't expect her to have to use it more than once.

But as of 1998, if you picked a random woman walking into an abortion clinic, there was almost a 50-50 shot she'd already had one. The number hasn't changed much. And as I've previously noted, a woman will have an average of .9 abortions before she turns 45, though only 1 in 3 women aborts even a single child by that age. Combine those two facts, and there are almost three abortions for each woman who aborts at all.

If a fetus is not a human life, this presents no problem whatsoever. But this nuanced position -- abortion is bad, and no one wants to have one, but it's only done with careful consideration, so we should keep it legal -- doesn't work with the facts. Too many abortions take place with women who've had at least one chance to learn.

Don't get MADD, get a machine

Stephen Dubner is back with an interesting story, but some lightweight analysis:

[There is] a machine you can put in a bar or restaurant that lets you measure your blood alcohol level so you know if you're fit to drive or not. This takes the guesswork out of a guessing game that has been going on for generations. It has been introduced in the U.S. by Federico Forero, a longtime manager of a probation office in Atlanta, who saw the machine in use in bars in Europe.

So far so good, but:

Also, the way it's configured, a customer has to pay to use the machine — that's how the bar owner is supposed to make his money back, by collecting fees for use. But will people be willing to pay, even a small fee, to find out news that they might not want to know? Would you weigh yourself in the bathroom every morning if your home scale cost $.50?

Uhhhh... bathroom scales do cost money, and people pay for them. Also, if the "news that they might not want to know" is "you can get out of a huge ticket, or even a fatal accident, by not driving right now," people would probably pay 50 cents for it. (If that is indeed the cost for a use.)

Also, one time I was visiting home in Wisconsin, one of my friends told me about a bar there with a Breathalyzer. Not sure how many lives it saved, but he said some people took it as a competition. So that's another concern.

More on the not-so-fair tax

Awhile back, on Jeremy Lott's blog, I wrote about the revelation that the Fair Tax (a flat national sales tax to replace the current income tax) would actually be regressive. Poorer people spend a higher proportion of the money they earn, so a consumption tax affects them disproportionately.

Throw in a monthly "prebate," and the tax becomes progressive relative to consumption (those who consume more pay a higher percent of consumption in taxes), but not necessarily relative to income.

Neal Boortz, one of FT's biggest advocates, keeps denying basic math:

By way of example, using current poverty statistics the "prebate" for a household of four people would be $506.00 per month. Add that $506.00 to the fact that no household will see anything deducted from their checks for income taxes or for Social Security or Medicare taxes … and you see a substantial rise in real income ... The president's own tax reform commission stated that the FairTax was the only tax reform plan out there that would completely untax the poor (at the federal level).

Actually, the poor are already pretty much untaxed at the federal level -- in fact, due to the Earned Income Tax Credit and other breaks, people in the $0-$15,000 and $15-$30,000 categories, on average, get back more than they paid. One third of income tax returns are from people with no tax liability, and a third of those even worked full-time all year.

Also, the commission Boortz mentions dedicated chapter nine of its report to the proposal and did not recommend FT. Its very first objection was that "absent a way to ease the burden of the retail sales tax on lower and middle-income Americans, [it] would not meet the requirement ... [of being] appropriately progressive."

Its second was, "Although a program could be designed to reduce the burden of a retail sales tax on lower-income and middle-income taxpayers by providing cash grants, such cash grants would represent a new entitlement program -- by far the largest in American history. [This] would cost approximately $600 billion to $780 billion per year and make most American families dependent on monthly checks from the federal government..."

Even under such a Fair Tax, most people would pay more than they do now. The report's Figure 9.4 looks at the difference between today's tax law and a "full replacement retail sales tax proposal with prebate by income level" (which may differ in the particulars from Boortz's proposal). The FT is better only if you're in the $0-$15,000 or $200,000+ categories. Looking at the data in terms of deciles, the lowest-earning 20 percent and highest-earning 10 percent of Americans would benefit.

Those earning beneath that range get more back in the "prebate" than they would in the current Earned Income Tax Credit (because they're not earning much income), and those making more than that spend low percentages of their incomes, thus paying little sales tax.

The report noted other concerns as well. For example, when you collect all your tax in one place, you need a huge tax, and a huge tax is worth a lot of effort to evade. Between enforcement and constantly sending out checks, the federal government could spend as much administering this system as they do the current one.

**

On to the second part of Boortz's article, a response to Rudy Giuliani. Giuliani opposed FT because, in today's real estate market, people need to be able to deduct mortgage costs. Deductions happen on income taxes, and there's no mechanism in place for deductions off the Fair Tax (in theory, you could set it up so people apply for bigger prebates, but that further diminishes FT's simplicity).

Boortz:

[L]et's approach one of these home owners and ask them a simple question. Would you rather have a nice income tax deduction so that your taxes would be lower, or would you not have to pay income taxes at all? The point here is so ridiculously easy even a politician can understand it --- a tax deduction is of no value whatsoever to someone who does not owe taxes! Under the FairTax all income taxes are gone! You are never going to find someone whining that they wish they owed some taxes so that they could make use of a tax deduction.

What Boortz misses, of course, is that no matter which system we use, the government will get its money somehow. If it gets it through an income tax, you can deduct your mortgage. If it gets it some other way, you can't. Therefore, the income tax is friendlier to people who use big deductions (and correspondingly less friendly for people who don't).

The better question to ask this homebuyer is: Say you could pay $5,000 in sales tax, or you could owe $5,200 in income tax but use a mortgage deduction to save $500. In the latter scenario, the money you don't pay is made up by people who don't have deductions, or who earn much more income. Which would you choose?

(For the record, I'm anti-deductions. They are regressive . We should replace them with tax credits or eliminate them altogether. Read about the difference here.)

I like parts of the Fair Tax -- it rewards earning and saving, and discourages spending. A limited national sales tax, coupled with lowering other taxes, might work to even out incentives. But it's just not good enough to replace virtually every tax we have.

Stephen Metcalf: Bringing us back to 1997

...in his new article about race, genetics and IQ. He spends the whole time playing statistics games, trying to find studies that control for this or that and conclude the IQ gap is completely environmental. He finds problems with studies that give different results.

Sorry, but that method of analysis is so last decade. Current research is working to locate specific genes that increase and decrease IQ, and then looking to see how those genes differ by population. Half Sigma has been documenting the studies, and while the results aren't in completely in yet, it's now an absolute fact that some of the pertinent genes vary by race. For example, William Saletan, in the article series Metcalf is responding to, pointed out the head-size genes. Metcalf doesn't even bring this up.

Metcalf also ( again) displays his outright ignorance of Arthur Jensen, one of the most important researchers on this issue:

Arthur Jensen has spent the last 40 years arguing against "compensatory education," or the idea that programs like Head Start have any efficacy in alleviating black underachievement.

Actually, no. As Thomas Sowell (who's on Metcalf's side of the debate) wrote:

Unlike others on the heredity side of the argument, Professor Jensen saw no need to dismiss environmental factors or to claim that some races were fit only to be hewers of wood and drawers of water.

One of the ironies of Jensen's landmark article was that it argued that the educational performances of children from disadvantaged groups could be greatly improved, even if there was no corresponding improvement in IQ scores.

There's a difference between criticizing Head Start and claiming that no educational technique can narrow the achievement gap -- everyone agrees that the gap is at least part environmental, including Jensen.

And then this load of hogwash:

Does it feel as though researchers like Jensen and Rushton, the so-called "race realists," have spent their careers examining a range of competing hypotheses for the black-white IQ gap, and carefully scrutinizing the quality of the research at their disposal? Or have they been attempting, at all costs, to prove a single hypothesis—that blacks are congenitally dumber than whites? Shouldn't researchers on any highly charged subject be required to show a minimum of clean hands?

Rushton is debatable, but Jensen most definitely has been researching with an open mind. He discovered the "environmental cumulative effect" when he observed that, in rural Georgia, older black siblings have lower IQs, and this doesn't happen with whites. The more children are subjected to bad environments, the lower their IQs get, and this plays a role in the race gap.

And back to Sowell:

Professor Jensen pointed out back in 1969 that black children's IQ scores rose by 8 to 10 points after he met with them informally in a play room and then tested them again after they were more relaxed around him. He did this because "I felt these children were really brighter than their IQ would indicate." What a shame that others seem to have less confidence in black children than Professor Jensen has had.

Also, again, no one supports a "heredity-only thesis"; the only debate is whether the gap is partially or completely environmental.

Steve Sailer has more here.

A self-tuning guitar?

It'll probably be 10 years before this becomes standard -- if it ever does -- but it could really make guitar playing a lot more fun. I hate tuning, especially when I just want to play a few notes in between other tasks. Also, in college I played in a band that did "You Shook Me All Night Long," and one time my G string slipped a bit flat before we started (for non-players: that's really, really bad).

It could also spark a resurgence of alternate tunings. I've messed around with them at various points in my evolution as a player, but once you've tuned your guitar to (as I once did) Eb G Db G Bb F, you can't play your stock riffs until you put it back into "normal" mode. This would make the process a breeze.

Instruction manual here.

Um, OK

Supergenius Stephen Dubner of the Freakonomics blog writes:

What I don't understand is why Imus got fired for his sins, albeit temporarily, while Family Guy rolls merrily along. I am not saying that Family Guy should be canned, or that Imus shouldn't have been, but it's a pretty curious situation.

He's not saying there's any reason anything should have gone differently. It's just curious it didn't!

He goes on to admit never watching Family Guy, and then demonstrating same for good measure, including an assertion that the dog is homosexual -- actually, the male dog dates human women, and it's the infant around whom the connotations fly.

He offers up these reasons for the "curious" phenomenon that shouldn't be any different:

1. Imus is human and Family Guy is a cartoon.

2. Imus is non-fiction and Family Guy is fiction (although it often has non-fiction elements).

3. Imus aspires to some level of intellectual sophistication while Family Guy is brazenly juvenile.

4. Imus is live talk while Family Guy is taped entertainment.

5. There is no real difference between the two, but the kind of big public storm that resulted in Imus being fired is essentially a random event, unpredictable and nearly inexplicable, and it typically arises when political, social, and media pressures all align just right. It can't be concocted, or controlled. It happened to Imus because it happened; and it hasn't happened to Family Guy just because it hasn't.

Well, for starters, humans write Family Guy. And if anything, taping something offensive, mulling it over and choosing to release it is worse than an off-the-cuff remark.

How's this: Imus singled out specific people who were not legitimate targets (the basketball team), said derogatory things about them and did so in his own voice. Family Guy makes fun of celebrities and wide, indiscriminate groups of people (say, whites or blacks in general) and does so through characters who behave inappropriately. There's nothing curious about it.

Here come the health-care stormtroopers

Slate quotes John Edwards:

[My plan] will require proof of insurance when income taxes are paid and when health care is provided. Families without insurance will be enrolled in Medicare, Medicaid, S-Chip or another targeted plan or be assigned a plan within new Health Care Markets. [These are regional markets Edwards would create in which private insurers would be invited to compete with public health care plans.]

Families who lose coverage will be expected to enroll in another plan or be assigned one. For the few people who refuse to pay, the government will help collect back premiums with interest and collection costs by using tools like the ones it uses for student loans and taxes, including collection agencies and wage garnishment.

I really don't have a dog in the health-care fight -- my proposal would be universal vouchers, and no one seems to support that -- but Edwards doesn't sound all that American here. It's the ultimate government intrusion into health decisions.

Second Amendment piece up at The American Spectator

Here it is. It's a response to The Harvard Crimson's editorial arguing for the Second Amendment's repeal.

Main point:

It's rife with problems, but the editors get more right than most of the gun-control movement does.

When it comes to guns and the Constitution, there are three main areas of importance: Historical (in what context was the Second Amendment written, and how has that context shifted?), legal (precisely what forms of gun control does the amendment proscribe?) and practical (does the Second Amendment prohibit more good policies than bad, and therefore, should we repeal it?). All told, the paper is about half-right.

They get half a cheer for their history (the context in which the Second Amendment passed has changed, as they argue, but they get the particulars wrong), a full cheer for their law (to pass strict gun control, we'd need to repeal the Second Amendment) and no cheer for their practical arguments (there's no case that the Second Amendment does more harm than good).

Sheesh

The poor media keeps reporting horrible bias incidents, then having to follow up with stories like this:

A firefighter who reported finding a knotted rope and a threatening note with a drawing of a noose in an East Baltimore station house last month had placed the items there himself, city officials said yesterday.

Brief review of Resident Evil: The Umbrella Chronicles

I finally finished the Wii game on normal difficulty today (except the ridiculous gallery minigame you unlock), and I have to say, on the whole, I found it very enjoyable. Addictive, even.

A few observations:

--I took about 20 hours. Each stage you beat earns you stars (even if you re-play stages you've already finished), which you can use to upgrade your weapons. The 20 hours includes re-plays of beaten stages, as well as multiple tries at first-time finishes.

--The graphics are OK, but nothing spectacular like those on RE4.

--In the early stages especially, the music is pretty terrible. Techno is not scary. In fact, one the whole, the game is more arcade shooting than horror -- there are some scenes that can creep you out a bit, but nothing makes the heart beat fast like much of RE4 (and the other RE titles) did.

--There's a surprising amount of strategy for a point-and-shoot light-gun-style game. You have to conserve ammo, decide when to pick up health when you have multiple chances to, make split-second decisions for dealing with enemies getting ready to hit you, etc. (Shooting them in the legs works well.) The bosses, while frustrating -- I should tally the swear words it takes to beat each game I play -- force you to develop strategies for dealing with each attack. Once you figure it out, you can beat them without taking very much damage at all.

--Many reviewers have decried the difficulty of getting "critical" or one-shot-kill hits; you have to shoot the zombie at the top of its head. It's true enough, and this makes it hard to get a good ranking on some stages (you're awarded stars based on whether you got an S, A, B or C), but it's very possible to beat any stage without a single critical hit. Once you get a knack for them (maybe 30-50 percent success?), they work well for conserving ammo -- the pistol has unlimited bullets, so if it can instantly pop zombies' skulls, you can use your heavy-duty stuff on a boss instead.

--It's kind of stupid that the boss of the game is in a hidden stage (get an A or S ranking on Raccoon's Destruction 3, then beat the stage that unlocks to get to Dark Legacy 2). It's also weird that beating him unlocks yet another stage.

--There are some frustrating glitches that really should have been fixed. For one, when you "retry" a stage, you can't pick a different weapon, so once you decide on a gun, you're stuck beating the whole stage with it, even if it turns out to be a bad match. Similarly, once you start a stage, you can't abort and try a different one instead, at least without going through a screen that implies it will delete your entire game. Maybe I'm misreading, but I didn't want to risk it.

--The health system is too primitive. You get a health bar, and when it's depleted, you're dead. You can pick up herbs that replenish more than half of it, and you can get first aid sprays that refill the whole thing when it runs out. The two don't interact well -- essentially, the game creates a needless dichotomy around whether you get to an herb before you need your first aid spray. Say a health bar is equivalent to 50 damage (the game assigns no numbers) and an herb refills 35. I have a first aid spray and a full bar (100 total health), and I have to go through a bunch of enemies to get to an herb. If I take 51 damage from those enemies, the first aid spray gets used and I'm down to 49 total health. When I get the herb it hardly does anything, because you can't fill your bar beyond 50. However, if I take 49 damage, the herb will increase my health from 1 to 36, and I'll still have the first aid spray for 86 total health. Taking 2 extra damage costs me 36.

All told, I'd give it 4/5 stars.

UPDATE: I've been getting some Google hits asking how to use the first aid spray. You don't use it; it automatically re-fills your bar when you run out of life.