Alpha Rolls, Treats, and the Dog-Cat Relations

In my previous post I admitted that Lori alpha rolled CC a few times and that it had a positive impact on her behavior. Looking back on that, I must admit I have some mixed feelings. My favorite dog-training book is "The Other End of the Leash" by Dr. Patricia McConnell, and I am sure she woudn't approve of an alpha roll or anything else you might see on the "Dog Whisperer."

Did it work? I think it's hard to say, but two things are clear:

• The immediate effect was to modify CC's behavior in a significant and positive way - we saw reduced jumping and much more response to command instruction (that is, when Lori asked CC to sit, she was more likely to sit).
  
• I don't believe it will have any long term effect -- I think the only value of what happened is that perhaps it created a window of more subdued behavior during which we can use extinguishing and lure-reward (that is, ignoring CC when she jumps and treating her when she sits).
• The situation was our fault - lure-reward and extinguishing could have been adequate had we acted much sooner. The reason we couldn't apply it in the more extreme case was logistical: CC is big enough and has sharp enough claws that if we ignore her and she persists in jumping, we can get scratched pretty badly.

If you look up Cesar Milan on Wikipedia you'll find that a lot of heavy-weight academic behaviorists are pretty unhappy about what he glamorizes on national television. I don't agree with all of their criticism, but I do agree with one: what he does is complex and very sensitive to timing, amount of force, and very careful reading of the dog's behavior. If you use lure-reward training, the worst thing that happens is your dog gets fat. But try what Cesar Milan does without his particular natural aptitude for dogs and you're likely to make the dog a lot worse, and possibly get injured in the process. I can only hope there aren't too many "amateur Cesars" out there trying to mimic bites with their hands and issuing corrections that are too harsh, at the wrong time, and for the wrong reasons.

(On the other hand, I think that the Dog Whisperer does contain some reasonable advice - increasing the time of CC's walks has done more to curb bad behavior than anything else, and we have realized how much our mental state and body language affect her behavior.)

As for the dog and cat, the situation continues to be problematic. Nala is quite nervous about CC and is easily scared off by a bark. Furthermore, she prefers to try to get by the dog when we are not watching, which means that a dog-cat chase tends to break out just when we think that we don't need to be paying attention. (Turns out that we need to be paying attention all the time.)

To try to improve things, we moved the dog's crate away from the door to the room with the cat's litter box, so the cat wouldn't have to approach the dog so closely to get in. This seems to be helping - the cat goes down there more often and doesn't seem to hesitate, and the dog barks at her less.

CC will respond to a verbal correction or a command -- if I am trying to keep the dog away from the cat what seems to work best includes:

• Early detection. If I catch CC watching the cat and give her alternate instructions, she will virtually always comply. Once she is up and going after the cat, she tunes us out.
• Blocking - it's not necessary to touch the dog to get her not to chase the cat. Putting a leg into the space she wants to go to, blocking her, will cause her to stop motion in that direction.
• Stay - she'll hold a stay (in response to a flat "stop" hand out) and keep her eyes following me if requested.
• I've actually found the "Cesar Milan ssh" noise works really well - it doesn't make her more stressed, or raise her excitement, anxiety or aggression level, but she always notices it and refocuses on me. The noise Cesar Milan makes strikes me as consistent with Dr. McConnell's research on the effect of frequency and duration on animal behavior. "ssssh" definitely gets better results than "no", I think because of the difference in pitch.

My hope in the long term is that with enough practice (I'll sit on the stairs and feed the cat while dog watches - the dog gets a treat if she stays calm during the exercise) I can break the cycle of dog-chases cat.

Bob Vila would not chase the cat

We just finished our second week of puppy-kindergarten with CC. It's been a transitional time for her behavior, and what we see in class illustrates it pretty well.

When it comes to demonstrating the things she's "supposed" to be learning (sit, lie down, watch me, come here, etc.) CC is the star. If the trainer needs someone to demo a trick, she's good for it. She is very motivated (especially for treats) and a very fast learner. She had learned most of her homework after one day last week.

But...when you stop working with her and ignore her, that's when things go down-hill rapidly. She'll bark and jump up and generally instigate until you pay attention to her. Looking back, it's clear to me that these behaviors are my fault - I taught them to her. When we first got her, for the first few weeks, her behavior was just totally unmanageable. She wasn't listening to us at all, so we felt like we had no control of the situation. I got fed up with it and decided that, no matter how long it took, I would teach her to reliably sit, lie down, and stay.

I work at home, so I have access to her all the time, and it turns out that "how long" was only a few days. She's really smart and figures out what you want her to do for a reward in no time. But, with such success with "training behaviors", I would use training as a way to keep her busy when she was bored, and that evolved into training her when she "asked" to be trained by barking, jumping, and generally being bad until we paid attention to her. So essentially I trained her to bug us for attention.

I'd been struggling to understand this last week - one of the reasons I think the behavior wasn't more clear is that this is also the week that Lori decided to deal with domination-related jumping. When Lori first got CC, her reaction was to give her a big hug and go "I love you so much" and crawl around on the ground with her. If you watch the Dog Whisperer you can see how this can amplify behavioral problems by a combination of not establishing dominance (e.g. letting the dog jump on Lori), rewarding bad behavior, and exciting the dog all at once.

This week Lori tried something different - she came home and CC ran to her. Lori said sit and the CC ignored her and jumped up. At that point instead of just backing away, Lori scruffed her and pinned her down. After a few applications, CC has stopped the dominance jumping (she'll still jump up to get attention, but I think you can tell by how she jumps what her intention is) and obeys commands much more readily when Lori issues them.

In the process of all of this, CC had a period of trying to dominate me. I'm not sure what her dog-logic was, exactly, perhaps something like "I can't be number 3, I have to be at leaset number 2 in the house". So intermingled with bad behavior for attention we've also had a reorganization of our pack, which led to some confusing behaviors. I think we're past the worst of the dominance behavior though, so we'll see how the next week goes.

I went to the vet to get tutored!

While Lori and I were on vacation we boarded CC at the local vet, and neutered her. She got her stitches out today. (And despite all of this vet-related trauma she still loves the car and thinks the vet is a dog park. She has a very optimistic personality.)

Just when her fur had grown back, they shaved it off again!

"What? You brought the dog back??"

This is how I learned that an unsupervised dog who likes the dig, left in the back yard with the sprinkler on, is not a good idea.

We had started to crate-train CC before her trip to the vet, but after a week of boarding she was pretty much used to it. This works out well for us...the sun-room isn't air conditioned, so we didn't have a good place to keep her when we wanted to go out and it was hot out.

"I'm in jail!"

Step 1 - Cut a Hole in the Wall

Ami came down to visit last week, which means housing destruction. We installed a pass-through panel. Basically this is a set of plugs for professional audio and Ethernet that run from one side of the wall to the other. This will allow me to close my studio door while recording audio in the rest of the house (the hard wood floor makes a nice sound room) and also pass Ethernet from the cable model in the living room to the computers in the office.


Here's the hole we cut in the drywall. A view of the living room.



Here is the panel's back, and Ami wiring it. The office is in total chaos - CC rewired it slightly, but that'll be a story for another post.

The final result.

Raining Cats And Dogs

Life with cat and dog remains a bit of a three-ring circus.

CC likes to hang out in the family room. She'll play with a rope or another toy while we hang out, but unfortunately her favorite toy is, well, Nala.

Nala likes to sleep on the chair in the family room and refuses to give up territory (probably good). Even though we've made a few cat-only rooms (using hooks to keep the door cracked) the cat will follow the dog around, watching from the other side of the room, until the dog comes over to play, at which point Nala hisses and swats and acts in a quite un-lady-like manner. (We got that cat tree to give Nala more high places to stay, but she just stays in the ground and instigates.)

CC is a little bigger than we planned for a dog - 40 lbs and growing. She can hold a regulation frisbee in her mouth easily, but doesn't really know how to play games with it. We're working on the basics first (stay, down, come, don't torment the cat, even if she started it).

A rare moment of peace in the living room - everybody's sleeping!

It's a cat in a box!

Puppy Puppy Puppy Puppy Puppy Puppy

Meet Cecelia!



She's a 7 month old, um...dog - she looks like a Golden Retriever to me, but the family that helped us with the adoption think maybe there is some Border Collie in her.



She loves to play and has boundless energy - if we could harness puppy power we wouldn't need foreign oil.



Lori helping her find dinner.



Nala is taking the situation pretty well, considering that her home has been invaded by a 30 lb 7 month old dog. We're keeping them separated and CC on a leash while they meet each other.



A meeting of minds. We bought a child-proofing gate - once we're letting CC roam the house we'll partition areas off so the cat can have a sanctuary.



Why am I in kitty jail? (Actually there are a dozen ways Nala can get over, under and around the fence.)

Curtains and Bushes

Lori's parents came down to DC this week for Zemer Chai's "Shalom y Esperanza" concert. Whenever they visit, serious work gets done on the house.


My mother made us curtains last fall as part of our redoing the living room. Finally this weekend we put them up!


Lori's parents also did a lot of work in the garden...unfortunately the bushes in front of the house died in the hot dry weather last summer. We bought two Azalia's and two Japanese Holly and put them in. That monster next to the car is the left over stump from a dead bush that we had to dig out. It's about a foot in diameter and ways more than you can imagine.


We also put some herbs into the backyard garden. Unlike our last herb garden (which actually did pretty well), we're watering it. We have rabbits in the back yard so there may not be much left for us.

Studio "Gear" I Like

I am a big fan of the virtualization of recording studios...that is, more and more the effects/sounds we had to use thing for are now available via software. I don't have a lot of boxes in my project studio, and having moved it 4 times now I like it that way.

I am not advocating for the virtualization of instruments...I think the feel and response of a
real instrument is necessary for the musician to create a real performance. I use drum machines because it's what I have, not because I think they could ever come close to the performance of a real drummer. But when it comes to processing gear, I think software is great. Whatever the real box does, you can create that in software, and doing so is driving down the cost of processing such that hobbyists like myself can have processing that would have only been found in big expensive studios a decade or two ago.

With that in mind, my favorite plugin (audio processing software for ProTools) is Trash. Trash destroys sound. Trash is a simulator not for whole guitar amps, but for all of the parts of a guitar amp, each individually tweakable, and it just sounds great. Back at WBRU, Steve and I used to use an old beat-up guitar amp to make a sound effect we just couldn't get any other way. Trash is the first distortion/amp modeler I've played with that can capture that, well, whatever it is.

Trash also has a wonderful feature that you don't find in real amps: it can act separately on three frequency bands. For example, when processing drums you can use different settings on the kick, snare, and hihat, each of which tends to live in its own frequency area. This lets you have that fat destroyed snare sound, crispy fried simple sound, and still leave the kick drum relatively unprocessed (so it doesn't get muddy). On vocals you can fuzz out the midrange but keep the high-end a little bit clean so you can hear what's going on.

I just said a drum machine can't replace a real drummer. Absolutely true. But drum machines are also their own things...I think electronica gets a hypnotic sound from the use of drum machines, loops, and computers. So sometimes I want a drum machine, and Rebirth is my favorite. Sadly Rebirth is out of production (sad because it was never ported to OS X, so while it will run on a PPC Mac under Classic or on Windows, its days for the Mac community are numbered).

Propellerhead Software did something that I think is really great: they posted Rebirth for free...a lot of software that is end-of-lifed simply ends up lost...the company puts the master code on a DVD-ROM for backup, and then the engineers who all worked on it move on to other jobs and eventually no one knows (1) where the code is or (2) how to build it. What a waste...to me the best thign would have been if Propellerhead had posted the source code (I would gladly port ReBirth to OS X for free), but it's their software and we're lucky just to have the CD-ROM image.

Rebirth is a simulation of the 808, 909 and 303 drum machines, plus you can plug in free sound modules. To hear what Rebirth sounds like (oh lord, another plug) take a listen to track 3 - "Min Hameitzar" from Ami's album here. That drum loop is the Rebirth 909 simulation with a plug-in module (Orbit, I think but I'm not sure.)

Even more fun is Rebirth + Trash + Beat Detective. (Beat Detective is a ProTools feature that lets you cut up drum loops and apply "feel" to them, moving the beats a little bit.) Start with Rebirth, program a loop, use Beat Detective to loosen it up a bit, then Trash to beat the hell out of it...I wouldn't describe the result as comparable to a real drummer, but it's not a bad sound.

Shameless Plug

We're going down to Norfolk this weekend for an Amichai Margolis Band gig...there's some audio on Ami's myspace page.

The Return of the Studio

Progress! With the floors down we could finally start moving furniture into the living room and unpack all of our stuff. For about six months we've had all of our books, games and display items still in boxes in a huge pile. Here are some pics of the living room and family room starting to take shape.



That area witih CDs and DVDs below the front door used to be entirely buried in boxes.

About half of our packed stuff was in the family room and the other half was in the back of my studio (the part you don't see in any of the pictures). Unfortunately which half was where was pretty random. So by the time we got the books unpacked, my studio looked like this:



I then have had time to clear that mess out too - the studio is almost entirely accessible.



So much more floor space is freed up, it's like having a whole second room.

Of course, no blog post would be complete without the obligatory cute cat picture. Nala only stands on all-tube preamps.

One DIY Book

I feel good about the work Lori and I have done on the house, as well as the work we've contracted out. I also feel that doing the work ourselves is in some cases a better value proposition than buying an end-product by paying more for a house. I would never extend this logic to infrastructure problems - only interior.

When does it pay to hire experts? When is DIY the right approach? When it comes to retirement and financial planning, I am way in the DIY camp. So I thoroughly enjoyed canceling my Bank of America "Premier" status today.

Premier status? Well, it turns out that somehow through a series of mergers and acquisitions, I got lumped in with Bank of America's rich clients. It turns out I had a "relationship manager" (whom I have never spoken with until today to cancel the whole thing). Since I don't have gobs of money and don't need gobs of advice, I don't think I need my relationship to the bank to be managed. Every time I have compared BoA to other institutions, they've proven uncompetitive. A professional that can "advise" me to buy these products simply illustrates the enormous conflict of interest present in financial services: it's a zero sum game between you and your banker/broker/retirement planner/financial services company. They make all their money by taking some of your money.

Can you do better on your own? I think so. There is an enormous amount written about investing and retirement planning. If you only read one book on the subject (perhaps to see if you can stand it) I suggest "A Random Walk Down Wallstreet" by Burton G. Malkiel. Four things going for it:

• It contains real investment information that you can use to set up your own portfolio and take action.
• It covers the academic topics behind investing, with clear explanations. This gives you the background to refute the next salesman who says "ah, but you can beat the market with our new XYZ mutual fund."
• The book has been updated recently, important both for market perspective and for covering new investment vehicles.
• Most importantly, it's a good read and goes quickly. I think the biggest barrier of entry to DIY finance can be the subject matter.
  

(John Bogle has probably done more for individual investors than anyone else I can think of, and he's very smart...I have the utmost respect for him. But his books can be dry...if you were to pick up Common Sense on Mutual Funds as a first book on the subject, it could also be your last.)

The original purpose of this blog (now it serves as a photo album) was to document my DIY mistakes on the house. But I figure financial DIY mistakes should be fair game too!

The Walls Are Not Purple!

Someone mentioned that our lower living room walls look purple. They are not! This is an artifact of the poorly-taken digital photo I posted, with no color retouching.

The real color is "milk chocolate" (Behr color 710B-5) and it does have a purple undertone, but it really looks mostly brown in real life...the flash just lit it up funny!

Bob Vila Would Not Invest

I suppose it was only a matter of time before I started ranting on the blog about financial matters. Last week I examined Vanguard's "lifecycle" funds, one of Lori's options for her 401k. I like these funds, but the comparable idea at Fidelity ("Freedom" funds) are just awful...the devil is in the details.

The idea of these "fund of funds" is rooted in Modern Portfolio Theory (MPT). Basically:

• If markets are efficient, you only get paid to take risk.
• Furthermore, you only get paid to take risk that you can't diversify away.

When you add these together, it becomes clear that the true controller of your portfolio's performance is what mix of asset classes you have, not what specific assets you have.

I am not saying that you can buy just any stock...we have to consider how diversification works. If a market is truly efficient, the expected return on any one stock should be the same as any other. (If it wasn't, smart people would buy the undervalued stock, driving its price up until the expected return matched.) If we think Coke is solid and American Home Mortgage is screwed, we should see a big difference in price that exactly equalizes their differing situations.

But that doesn't mean that investing in either one is the same. On average, they should return the same, but if you invest in only AHM there's a good chance that the company will go broke and you'll have nothing. The solution is of course to invest in both, or better yet, to invest in all stocks.

In other words, MPT says there is (to paraphrase Burton Malkiel) exactly one free lunch - diversification. We don't get paid extra to hold fewer stocks so we might as well hold a lot and cut down the fluctuations in our portfolio. (Thus what's going to matter is the ratio of stocks to bonds, not which stocks, because we should be holding them all.)

As a side note, when I told my friend Seth this, he said I was full of crap, because the highest return I could possibly get was in just one stock. This is true, sort of. The highest return I can get is in one stock, but not the highest expected value. The difference is between what will probably happen and what does happen. The problem with a concentrated portfolio (if you have one good stock idea, put all of your money into it) is that you have to know that that stock will do well. So if you're a "Mad Money" fan and believe that you can value stocks better than the market, it makes sense to concentrate. If you're an Efficient Market Theory nerd like me, you figure that your stock picks are probably average (or maybe worse than average) and therefore not worth concentrating in.

Vanguard

So MPT says we can trade off our risk and reward by changing what we invest in, and we should diversify whenever possible. The idea of the Lifecycle funds is pretty simple: you invest in the fund and the fund changes its ratio of assets over time to match a risk profile that's appropriate to your retirement date.

Consider the Vanguard 2040 retirement fund (VFORX). This is a retirement fund for people who still have on average 33 years in the workforce, so it's mostly invested in higher-risk, higher reward stock indices. Over time they'll start converting to bonds, lowering returns, but reducing the volatiliy of the fund. (That is, as you get closer to retirement, you'll expect to earn less, but the chances of your money being half-gone due to a market crash will go down.)

What I like about this fund is its low expenses. Mutual fund expenses are your money that the management company takes from you every year for the honor of fondling your money. How much you make is basically how much you theoretically would have made, minus what they pocketed. Costs are really tricky, because typically mutual funds do not charge you more if they make you more money. They simply charge you money all the time. Your interests and your fund's interests are not at all aligned.

I will comment more on expenses later, but there are two things to note about the Van Guard Lifecycle fund:

• The effective fees are 0.21% and
• It's invested entirely in index funds.
  

Fidelity

By comparison, I can only describe Fidelity's Freedom Fund 2040 (FFFFX) as a bit of a disaster.

• The fund is made up of a who's-who list of large, expensive and poorly performing stock funds. I know this because (blushes) I was invested in these. The top holdings were available to me for my 401k at a previous employer. I'll have to rant separately about these funds.
• There are so many stock funds that what we really have is "closet indexing". Basically, if you buy enough different stock funds, you've got so many stocks, and so littel of each one, that you basically get the average performance of the whole market. I don't believe that "Fidelity Growth and Income" is a better investment than the S&P 500 (hrm - it correlates with the market by 89% and yet earned 2% less on average every year) but even if you do think it's a good idea, it's only 10% of the Freedom Fund, so any unique value it has is being washed out by other funds.
• The expense ratio of this fund is 0.76%. In other words, we're paying almost a percent for the honor of holding a little of each of these train-wrecks.

To be clear: I'm saying diversification is good, and it's better to have a lot of different stocks than a few. But it's better to buy, say, the Spartan Total Market Return (FSTMX) index fund, expense ratio 0.10% than to glom together a ton of underperforming equity funds, expense ratio 0.76%.

The other side of this equation is bond funds. Fidelity has very cheap index funds (the "Spartan" index funds). Now it may be that these funds are cheap because Fidelity subsidizes them. (This happened with their international fund...Fidelity actually lost money on it as way to keep the expenses down. International funds are expensive even when indexed due to high transaction costs, etc.) My logic is: I don't care why it's cheap, as long as it is cheap. In the case of Fidelity vs. Vanguard, I would say vigilance is needed to make sure that Fidelity doesn't slowly sneak expenses back up.

But when it comes to bond funds, it's a different situation. Most of Van Guard's bond funds are around 0.20% expenses, while Fidelity's cheapest ones (the "Spartan" bond funds) are 0.45% expenses. I think perhaps the buying public hasn't put 2+2 together and realized that cheap indexed bond funds represent the highest available bond returns to investors.

(Expenses matter even more for bond funds than for stock funds - the actual "risk premium" - that is, the increased returns of the bond fund over a risk-free investment is just tiny, so those expense points eat up a lot more of what you're paying for - improved returns by holding a riskier investment. In other words, it's one thing to steal a point from a 10% return, but a different prospect to steal a point from a 6% return.)

Of course, that's all moot, because like the stock section, Fidelity's life-cycle fund is not invested in index funds, but in actively managed bond funds, with expense ratios that vary from 0.45% up to 0.75%.

Summary

In summary, the devil is in the details - life-cycle-style investing represent a reasonable service to investors in principle - a way to buy portfolio asset-class management in a cost effective way, by pooling it with a huge group of people who will all retire at the same age. But the actual value of these offers depends greatly on the costs of the vehicle you use.

A little math to sum it up: assume a very crude model: add $1000 a year to your 401k for 30 years, blending from 68% stocks at the beginning to 10% stocks at the end. We'll use 10% for the stock return and 6% for bonds. (I realize these numbers are arbitrary, but if they were lower, they would only make the point more dramatic. I think there is no risk of the market returning better than 10% on average for the next 30 years.)

The total value of your 30,000 investment should be about $99,068. In a fund with 0.21% expenses, it is reduced to $95,338 - a real drag. But in a fund with 0.76% expenses it is reduced to $86,287. That's around 8% of your money! (If we model this with more depressing returns, the fees play a larger roll and eat 12% of returns.)

The Floor is Done (well, almost)

The floor is in! We just need to put the T-transitions on the transitions to tile - the transition strips we have didn't fit due to the height of the tile. (Well, it's fake tile.)

Here are some pics:

Stairway to Heaven (or at least the upstairs closet)

Work on the floor continues. These two pictures illustrate an interesting problem with the stair nosings.

Our old subfloor had stair nosings built into the subfloor stair-wood. (This wood remains in place and the hardwood is nailed and glued to it to make the final stairs. When we bought the house rug was in place using tack strips.)

But the new wood requires a square corner (no nosing) so that the hardwood can be placed directly over it. The new nosing is "built into" the hard wood.

So our contractor cut off the nosings in the subfloor.

In this picture you can see the three stages. In the top step, the hole where the nosing fit into the molding has been filled in with extra wood. On the middle step, the riser is in place, covering part of this fill. On the bottom step, the nosing is in place too, completely obscuring the work.

Someone is not very happy about the noise.